diff options
Diffstat (limited to 'fs')
91 files changed, 12742 insertions, 1245 deletions
diff --git a/fs/Kconfig b/fs/Kconfig index 9d757673bf40..6725f59c18e6 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -209,6 +209,7 @@ menuconfig MISC_FILESYSTEMS if MISC_FILESYSTEMS +source "fs/orangefs/Kconfig" source "fs/adfs/Kconfig" source "fs/affs/Kconfig" source "fs/ecryptfs/Kconfig" diff --git a/fs/Makefile b/fs/Makefile index 252c96898a43..85b6e13b62d3 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -106,6 +106,7 @@ obj-$(CONFIG_AUTOFS4_FS) += autofs4/ obj-$(CONFIG_ADFS_FS) += adfs/ obj-$(CONFIG_FUSE_FS) += fuse/ obj-$(CONFIG_OVERLAY_FS) += overlayfs/ +obj-$(CONFIG_ORANGEFS_FS) += orangefs/ obj-$(CONFIG_UDF_FS) += udf/ obj-$(CONFIG_SUN_OPENPROMFS) += openpromfs/ obj-$(CONFIG_OMFS_FS) += omfs/ diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 4b02591b0301..d01f89d130e0 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -25,7 +25,6 @@ #include <linux/buffer_head.h> #include <linux/workqueue.h> #include <linux/kthread.h> -#include <linux/freezer.h> #include <linux/slab.h> #include <linux/migrate.h> #include <linux/ratelimit.h> @@ -303,7 +302,7 @@ static int csum_tree_block(struct btrfs_fs_info *fs_info, err = map_private_extent_buffer(buf, offset, 32, &kaddr, &map_start, &map_len); if (err) - return 1; + return err; cur_len = min(len, map_len - (offset - map_start)); crc = btrfs_csum_data(kaddr + offset - map_start, crc, cur_len); @@ -313,7 +312,7 @@ static int csum_tree_block(struct btrfs_fs_info *fs_info, if (csum_size > sizeof(inline_result)) { result = kzalloc(csum_size, GFP_NOFS); if (!result) - return 1; + return -ENOMEM; } else { result = (char *)&inline_result; } @@ -334,7 +333,7 @@ static int csum_tree_block(struct btrfs_fs_info *fs_info, val, found, btrfs_header_level(buf)); if (result != (char *)&inline_result) kfree(result); - return 1; + return -EUCLEAN; } } else { write_extent_buffer(buf, result, 0, csum_size); @@ -513,11 +512,21 @@ static int csum_dirty_buffer(struct btrfs_fs_info *fs_info, struct page *page) eb = (struct extent_buffer *)page->private; if (page != eb->pages[0]) return 0; + found_start = btrfs_header_bytenr(eb); - if (WARN_ON(found_start != start || !PageUptodate(page))) - return 0; - csum_tree_block(fs_info, eb, 0); - return 0; + /* + * Please do not consolidate these warnings into a single if. + * It is useful to know what went wrong. + */ + if (WARN_ON(found_start != start)) + return -EUCLEAN; + if (WARN_ON(!PageUptodate(page))) + return -EUCLEAN; + + ASSERT(memcmp_extent_buffer(eb, fs_info->fsid, + btrfs_header_fsid(), BTRFS_FSID_SIZE) == 0); + + return csum_tree_block(fs_info, eb, 0); } static int check_tree_block_fsid(struct btrfs_fs_info *fs_info, @@ -661,10 +670,8 @@ static int btree_readpage_end_io_hook(struct btrfs_io_bio *io_bio, eb, found_level); ret = csum_tree_block(fs_info, eb, 1); - if (ret) { - ret = -EIO; + if (ret) goto err; - } /* * If this is a leaf block and it is corrupt, set the corrupt bit so @@ -1831,7 +1838,7 @@ static int cleaner_kthread(void *arg) */ btrfs_delete_unused_bgs(root->fs_info); sleep: - if (!try_to_freeze() && !again) { + if (!again) { set_current_state(TASK_INTERRUPTIBLE); if (!kthread_should_stop()) schedule(); @@ -1921,14 +1928,12 @@ sleep: if (unlikely(test_bit(BTRFS_FS_STATE_ERROR, &root->fs_info->fs_state))) btrfs_cleanup_transaction(root); - if (!try_to_freeze()) { - set_current_state(TASK_INTERRUPTIBLE); - if (!kthread_should_stop() && - (!btrfs_transaction_blocked(root->fs_info) || - cannot_commit)) - schedule_timeout(delay); - __set_current_state(TASK_RUNNING); - } + set_current_state(TASK_INTERRUPTIBLE); + if (!kthread_should_stop() && + (!btrfs_transaction_blocked(root->fs_info) || + cannot_commit)) + schedule_timeout(delay); + __set_current_state(TASK_RUNNING); } while (!kthread_should_stop()); return 0; } diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 19adeb0ef82a..fc5cae2a0db2 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -175,8 +175,8 @@ static void ceph_invalidatepage(struct page *page, unsigned int offset, static int ceph_releasepage(struct page *page, gfp_t g) { - struct inode *inode = page->mapping ? page->mapping->host : NULL; - dout("%p releasepage %p idx %lu\n", inode, page, page->index); + dout("%p releasepage %p idx %lu\n", page->mapping->host, + page, page->index); WARN_ON(PageDirty(page)); /* Can we release the page from the cache? */ @@ -276,7 +276,7 @@ static void finish_read(struct ceph_osd_request *req, struct ceph_msg *msg) for (i = 0; i < num_pages; i++) { struct page *page = osd_data->pages[i]; - if (rc < 0 && rc != ENOENT) + if (rc < 0 && rc != -ENOENT) goto unlock; if (bytes < (int)PAGE_CACHE_SIZE) { /* zero (remainder of) page */ @@ -606,71 +606,71 @@ static void writepages_finish(struct ceph_osd_request *req, struct inode *inode = req->r_inode; struct ceph_inode_info *ci = ceph_inode(inode); struct ceph_osd_data *osd_data; - unsigned wrote; struct page *page; - int num_pages; - int i; + int num_pages, total_pages = 0; + int i, j; + int rc = req->r_result; struct ceph_snap_context *snapc = req->r_snapc; struct address_space *mapping = inode->i_mapping; - int rc = req->r_result; - u64 bytes = req->r_ops[0].extent.length; struct ceph_fs_client *fsc = ceph_inode_to_client(inode); - long writeback_stat; - unsigned issued = ceph_caps_issued(ci); + bool remove_page; - osd_data = osd_req_op_extent_osd_data(req, 0); - BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_PAGES); - num_pages = calc_pages_for((u64)osd_data->alignment, - (u64)osd_data->length); - if (rc >= 0) { - /* - * Assume we wrote the pages we originally sent. The - * osd might reply with fewer pages if our writeback - * raced with a truncation and was adjusted at the osd, - * so don't believe the reply. - */ - wrote = num_pages; - } else { - wrote = 0; + + dout("writepages_finish %p rc %d\n", inode, rc); + if (rc < 0) mapping_set_error(mapping, rc); - } - dout("writepages_finish %p rc %d bytes %llu wrote %d (pages)\n", - inode, rc, bytes, wrote); - /* clean all pages */ - for (i = 0; i < num_pages; i++) { - page = osd_data->pages[i]; - BUG_ON(!page); - WARN_ON(!PageUptodate(page)); + /* + * We lost the cache cap, need to truncate the page before + * it is unlocked, otherwise we'd truncate it later in the + * page truncation thread, possibly losing some data that + * raced its way in + */ + remove_page = !(ceph_caps_issued(ci) & + (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)); - writeback_stat = - atomic_long_dec_return(&fsc->writeback_count); - if (writeback_stat < - CONGESTION_OFF_THRESH(fsc->mount_options->congestion_kb)) - clear_bdi_congested(&fsc->backing_dev_info, - BLK_RW_ASYNC); + /* clean all pages */ + for (i = 0; i < req->r_num_ops; i++) { + if (req->r_ops[i].op != CEPH_OSD_OP_WRITE) + break; - ceph_put_snap_context(page_snap_context(page)); - page->private = 0; - ClearPagePrivate(page); - dout("unlocking %d %p\n", i, page); - end_page_writeback(page); + osd_data = osd_req_op_extent_osd_data(req, i); + BUG_ON(osd_data->type != CEPH_OSD_DATA_TYPE_PAGES); + num_pages = calc_pages_for((u64)osd_data->alignment, + (u64)osd_data->length); + total_pages += num_pages; + for (j = 0; j < num_pages; j++) { + page = osd_data->pages[j]; + BUG_ON(!page); + WARN_ON(!PageUptodate(page)); + + if (atomic_long_dec_return(&fsc->writeback_count) < + CONGESTION_OFF_THRESH( + fsc->mount_options->congestion_kb)) + clear_bdi_congested(&fsc->backing_dev_info, + BLK_RW_ASYNC); + + ceph_put_snap_context(page_snap_context(page)); + page->private = 0; + ClearPagePrivate(page); + dout("unlocking %p\n", page); + end_page_writeback(page); + + if (remove_page) + generic_error_remove_page(inode->i_mapping, + page); - /* - * We lost the cache cap, need to truncate the page before - * it is unlocked, otherwise we'd truncate it later in the - * page truncation thread, possibly losing some data that - * raced its way in - */ - if ((issued & (CEPH_CAP_FILE_CACHE|CEPH_CAP_FILE_LAZYIO)) == 0) - generic_error_remove_page(inode->i_mapping, page); + unlock_page(page); + } + dout("writepages_finish %p wrote %llu bytes cleaned %d pages\n", + inode, osd_data->length, rc >= 0 ? num_pages : 0); - unlock_page(page); + ceph_release_pages(osd_data->pages, num_pages); } - dout("%p wrote+cleaned %d pages\n", inode, wrote); - ceph_put_wrbuffer_cap_refs(ci, num_pages, snapc); - ceph_release_pages(osd_data->pages, num_pages); + ceph_put_wrbuffer_cap_refs(ci, total_pages, snapc); + + osd_data = osd_req_op_extent_osd_data(req, 0); if (osd_data->pages_from_pool) mempool_free(osd_data->pages, ceph_sb_to_client(inode->i_sb)->wb_pagevec_pool); @@ -778,17 +778,15 @@ retry: while (!done && index <= end) { unsigned i; int first; - pgoff_t next; - int pvec_pages, locked_pages; - struct page **pages = NULL; + pgoff_t strip_unit_end = 0; + int num_ops = 0, op_idx; + int pvec_pages, locked_pages = 0; + struct page **pages = NULL, **data_pages; mempool_t *pool = NULL; /* Becomes non-null if mempool used */ struct page *page; int want; - u64 offset, len; - long writeback_stat; + u64 offset = 0, len = 0; - next = 0; - locked_pages = 0; max_pages = max_pages_ever; get_more_pages: @@ -824,8 +822,8 @@ get_more_pages: unlock_page(page); break; } - if (next && (page->index != next)) { - dout("not consecutive %p\n", page); + if (strip_unit_end && (page->index > strip_unit_end)) { + dout("end of strip unit %p\n", page); unlock_page(page); break; } @@ -867,36 +865,31 @@ get_more_pages: /* * We have something to write. If this is * the first locked page this time through, - * allocate an osd request and a page array - * that it will use. + * calculate max possinle write size and + * allocate a page array */ if (locked_pages == 0) { - BUG_ON(pages); + u64 objnum; + u64 objoff; + /* prepare async write request */ offset = (u64)page_offset(page); len = wsize; - req = ceph_osdc_new_request(&fsc->client->osdc, - &ci->i_layout, vino, - offset, &len, 0, - do_sync ? 2 : 1, - CEPH_OSD_OP_WRITE, - CEPH_OSD_FLAG_WRITE | - CEPH_OSD_FLAG_ONDISK, - snapc, truncate_seq, - truncate_size, true); - if (IS_ERR(req)) { - rc = PTR_ERR(req); + + rc = ceph_calc_file_object_mapping(&ci->i_layout, + offset, len, + &objnum, &objoff, + &len); + if (rc < 0) { unlock_page(page); break; } - if (do_sync) - osd_req_op_init(req, 1, - CEPH_OSD_OP_STARTSYNC, 0); - - req->r_callback = writepages_finish; - req->r_inode = inode; + num_ops = 1 + do_sync; + strip_unit_end = page->index + + ((len - 1) >> PAGE_CACHE_SHIFT); + BUG_ON(pages); max_pages = calc_pages_for(0, (u64)len); pages = kmalloc(max_pages * sizeof (*pages), GFP_NOFS); @@ -905,6 +898,20 @@ get_more_pages: pages = mempool_alloc(pool, GFP_NOFS); BUG_ON(!pages); } + + len = 0; + } else if (page->index != + (offset + len) >> PAGE_CACHE_SHIFT) { + if (num_ops >= (pool ? CEPH_OSD_SLAB_OPS : + CEPH_OSD_MAX_OPS)) { + redirty_page_for_writepage(wbc, page); + unlock_page(page); + break; + } + + num_ops++; + offset = (u64)page_offset(page); + len = 0; } /* note position of first page in pvec */ @@ -913,18 +920,16 @@ get_more_pages: dout("%p will write page %p idx %lu\n", inode, page, page->index); - writeback_stat = - atomic_long_inc_return(&fsc->writeback_count); - if (writeback_stat > CONGESTION_ON_THRESH( + if (atomic_long_inc_return(&fsc->writeback_count) > + CONGESTION_ON_THRESH( fsc->mount_options->congestion_kb)) { set_bdi_congested(&fsc->backing_dev_info, BLK_RW_ASYNC); } - set_page_writeback(page); pages[locked_pages] = page; locked_pages++; - next = page->index + 1; + len += PAGE_CACHE_SIZE; } /* did we get anything? */ @@ -944,38 +949,119 @@ get_more_pages: /* shift unused pages over in the pvec... we * will need to release them below. */ for (j = i; j < pvec_pages; j++) { - dout(" pvec leftover page %p\n", - pvec.pages[j]); + dout(" pvec leftover page %p\n", pvec.pages[j]); pvec.pages[j-i+first] = pvec.pages[j]; } pvec.nr -= i-first; } - /* Format the osd request message and submit the write */ +new_request: offset = page_offset(pages[0]); - len = (u64)locked_pages << PAGE_CACHE_SHIFT; - if (snap_size == -1) { - len = min(len, (u64)i_size_read(inode) - offset); - /* writepages_finish() clears writeback pages - * according to the data length, so make sure - * data length covers all locked pages */ - len = max(len, 1 + - ((u64)(locked_pages - 1) << PAGE_CACHE_SHIFT)); - } else { - len = min(len, snap_size - offset); + len = wsize; + + req = ceph_osdc_new_request(&fsc->client->osdc, + &ci->i_layout, vino, + offset, &len, 0, num_ops, + CEPH_OSD_OP_WRITE, + CEPH_OSD_FLAG_WRITE | + CEPH_OSD_FLAG_ONDISK, + snapc, truncate_seq, + truncate_size, false); + if (IS_ERR(req)) { + req = ceph_osdc_new_request(&fsc->client->osdc, + &ci->i_layout, vino, + offset, &len, 0, + min(num_ops, + CEPH_OSD_SLAB_OPS), + CEPH_OSD_OP_WRITE, + CEPH_OSD_FLAG_WRITE | + CEPH_OSD_FLAG_ONDISK, + snapc, truncate_seq, + truncate_size, true); + BUG_ON(IS_ERR(req)); } - dout("writepages got %d pages at %llu~%llu\n", - locked_pages, offset, len); + BUG_ON(len < page_offset(pages[locked_pages - 1]) + + PAGE_CACHE_SIZE - offset); + + req->r_callback = writepages_finish; + req->r_inode = inode; - osd_req_op_extent_osd_data_pages(req, 0, pages, len, 0, + /* Format the osd request message and submit the write */ + len = 0; + data_pages = pages; + op_idx = 0; + for (i = 0; i < locked_pages; i++) { + u64 cur_offset = page_offset(pages[i]); + if (offset + len != cur_offset) { + if (op_idx + do_sync + 1 == req->r_num_ops) + break; + osd_req_op_extent_dup_last(req, op_idx, + cur_offset - offset); + dout("writepages got pages at %llu~%llu\n", + offset, len); + osd_req_op_extent_osd_data_pages(req, op_idx, + data_pages, len, 0, !!pool, false); + osd_req_op_extent_update(req, op_idx, len); - pages = NULL; /* request message now owns the pages array */ - pool = NULL; + len = 0; + offset = cur_offset; + data_pages = pages + i; + op_idx++; + } - /* Update the write op length in case we changed it */ + set_page_writeback(pages[i]); + len += PAGE_CACHE_SIZE; + } + + if (snap_size != -1) { + len = min(len, snap_size - offset); + } else if (i == locked_pages) { + /* writepages_finish() clears writeback pages + * according to the data length, so make sure + * data length covers all locked pages */ + u64 min_len = len + 1 - PAGE_CACHE_SIZE; + len = min(len, (u64)i_size_read(inode) - offset); + len = max(len, min_len); + } + dout("writepages got pages at %llu~%llu\n", offset, len); + + osd_req_op_extent_osd_data_pages(req, op_idx, data_pages, len, + 0, !!pool, false); + osd_req_op_extent_update(req, op_idx, len); - osd_req_op_extent_update(req, 0, len); + if (do_sync) { + op_idx++; + osd_req_op_init(req, op_idx, CEPH_OSD_OP_STARTSYNC, 0); + } + BUG_ON(op_idx + 1 != req->r_num_ops); + + pool = NULL; + if (i < locked_pages) { + BUG_ON(num_ops <= req->r_num_ops); + num_ops -= req->r_num_ops; + num_ops += do_sync; + locked_pages -= i; + + /* allocate new pages array for next request */ + data_pages = pages; + pages = kmalloc(locked_pages * sizeof (*pages), + GFP_NOFS); + if (!pages) { + pool = fsc->wb_pagevec_pool; + pages = mempool_alloc(pool, GFP_NOFS); + BUG_ON(!pages); + } + memcpy(pages, data_pages + i, + locked_pages * sizeof(*pages)); + memset(data_pages + i, 0, + locked_pages * sizeof(*pages)); + } else { + BUG_ON(num_ops != req->r_num_ops); + index = pages[i - 1]->index + 1; + /* request message now owns the pages array */ + pages = NULL; + } vino = ceph_vino(inode); ceph_osdc_build_request(req, offset, snapc, vino.snap, @@ -985,9 +1071,10 @@ get_more_pages: BUG_ON(rc); req = NULL; - /* continue? */ - index = next; - wbc->nr_to_write -= locked_pages; + wbc->nr_to_write -= i; + if (pages) + goto new_request; + if (wbc->nr_to_write <= 0) done = 1; @@ -1522,7 +1609,7 @@ int ceph_uninline_data(struct file *filp, struct page *locked_page) ceph_vino(inode), 0, &len, 0, 1, CEPH_OSD_OP_CREATE, CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE, - ceph_empty_snapc, 0, 0, false); + NULL, 0, 0, false); if (IS_ERR(req)) { err = PTR_ERR(req); goto out; @@ -1540,9 +1627,8 @@ int ceph_uninline_data(struct file *filp, struct page *locked_page) ceph_vino(inode), 0, &len, 1, 3, CEPH_OSD_OP_WRITE, CEPH_OSD_FLAG_ONDISK | CEPH_OSD_FLAG_WRITE, - ceph_empty_snapc, - ci->i_truncate_seq, ci->i_truncate_size, - false); + NULL, ci->i_truncate_seq, + ci->i_truncate_size, false); if (IS_ERR(req)) { err = PTR_ERR(req); goto out; @@ -1663,8 +1749,7 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, u32 pool) goto out; } - rd_req = ceph_osdc_alloc_request(&fsc->client->osdc, - ceph_empty_snapc, + rd_req = ceph_osdc_alloc_request(&fsc->client->osdc, NULL, 1, false, GFP_NOFS); if (!rd_req) { err = -ENOMEM; @@ -1678,8 +1763,7 @@ static int __ceph_pool_perm_get(struct ceph_inode_info *ci, u32 pool) "%llx.00000000", ci->i_vino.ino); rd_req->r_base_oid.name_len = strlen(rd_req->r_base_oid.name); - wr_req = ceph_osdc_alloc_request(&fsc->client->osdc, - ceph_empty_snapc, + wr_req = ceph_osdc_alloc_request(&fsc->client->osdc, NULL, 1, false, GFP_NOFS); if (!wr_req) { err = -ENOMEM; diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 6fe0ad26a7df..de17bb232ff8 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -991,7 +991,7 @@ static int send_cap_msg(struct ceph_mds_session *session, u32 seq, u64 flush_tid, u64 oldest_flush_tid, u32 issue_seq, u32 mseq, u64 size, u64 max_size, struct timespec *mtime, struct timespec *atime, - u64 time_warp_seq, + struct timespec *ctime, u64 time_warp_seq, kuid_t uid, kgid_t gid, umode_t mode, u64 xattr_version, struct ceph_buffer *xattrs_buf, @@ -1042,6 +1042,8 @@ static int send_cap_msg(struct ceph_mds_session *session, ceph_encode_timespec(&fc->mtime, mtime); if (atime) ceph_encode_timespec(&fc->atime, atime); + if (ctime) + ceph_encode_timespec(&fc->ctime, ctime); fc->time_warp_seq = cpu_to_le32(time_warp_seq); fc->uid = cpu_to_le32(from_kuid(&init_user_ns, uid)); @@ -1116,7 +1118,7 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, int held, revoking, dropping, keep; u64 seq, issue_seq, mseq, time_warp_seq, follows; u64 size, max_size; - struct timespec mtime, atime; + struct timespec mtime, atime, ctime; int wake = 0; umode_t mode; kuid_t uid; @@ -1180,6 +1182,7 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, ci->i_requested_max_size = max_size; mtime = inode->i_mtime; atime = inode->i_atime; + ctime = inode->i_ctime; time_warp_seq = ci->i_time_warp_seq; uid = inode->i_uid; gid = inode->i_gid; @@ -1198,7 +1201,7 @@ static int __send_cap(struct ceph_mds_client *mdsc, struct ceph_cap *cap, ret = send_cap_msg(session, ceph_vino(inode).ino, cap_id, op, keep, want, flushing, seq, flush_tid, oldest_flush_tid, issue_seq, mseq, - size, max_size, &mtime, &atime, time_warp_seq, + size, max_size, &mtime, &atime, &ctime, time_warp_seq, uid, gid, mode, xattr_version, xattr_blob, follows, inline_data); if (ret < 0) { @@ -1320,7 +1323,7 @@ retry: capsnap->dirty, 0, capsnap->flush_tid, 0, 0, mseq, capsnap->size, 0, &capsnap->mtime, &capsnap->atime, - capsnap->time_warp_seq, + &capsnap->ctime, capsnap->time_warp_seq, capsnap->uid, capsnap->gid, capsnap->mode, capsnap->xattr_version, capsnap->xattr_blob, capsnap->follows, capsnap->inline_data); diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index fd11fb231a2e..fadc243dfb28 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -38,7 +38,7 @@ int ceph_init_dentry(struct dentry *dentry) if (dentry->d_fsdata) return 0; - di = kmem_cache_alloc(ceph_dentry_cachep, GFP_KERNEL | __GFP_ZERO); + di = kmem_cache_zalloc(ceph_dentry_cachep, GFP_KERNEL); if (!di) return -ENOMEM; /* oh well */ @@ -68,23 +68,6 @@ out_unlock: return 0; } -struct inode *ceph_get_dentry_parent_inode(struct dentry *dentry) -{ - struct inode *inode = NULL; - - if (!dentry) - return NULL; - - spin_lock(&dentry->d_lock); - if (!IS_ROOT(dentry)) { - inode = d_inode(dentry->d_parent); - ihold(inode); - } - spin_unlock(&dentry->d_lock); - return inode; -} - - /* * for readdir, we encode the directory frag and offset within that * frag into f_pos. @@ -624,6 +607,7 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry, struct ceph_mds_client *mdsc = fsc->mdsc; struct ceph_mds_request *req; int op; + int mask; int err; dout("lookup %p dentry %p '%pd'\n", @@ -666,8 +650,12 @@ static struct dentry *ceph_lookup(struct inode *dir, struct dentry *dentry, return ERR_CAST(req); req->r_dentry = dget(dentry); req->r_num_caps = 2; - /* we only need inode linkage */ - req->r_args.getattr.mask = cpu_to_le32(CEPH_STAT_CAP_INODE); + + mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED; + if (ceph_security_xattr_wanted(dir)) + mask |= CEPH_CAP_XATTR_SHARED; + req->r_args.getattr.mask = cpu_to_le32(mask); + req->r_locked_dir = dir; err = ceph_mdsc_do_request(mdsc, NULL, req); err = ceph_handle_snapdir(req, dentry, err); @@ -1095,6 +1083,7 @@ static int dir_lease_is_valid(struct inode *dir, struct dentry *dentry) static int ceph_d_revalidate(struct dentry *dentry, unsigned int flags) { int valid = 0; + struct dentry *parent; struct inode *dir; if (flags & LOOKUP_RCU) @@ -1103,7 +1092,8 @@ static int ceph_d_revalidate(struct dentry *dentry, unsigned int flags) dout("d_revalidate %p '%pd' inode %p offset %lld\n", dentry, dentry, d_inode(dentry), ceph_dentry(dentry)->offset); - dir = ceph_get_dentry_parent_inode(dentry); + parent = dget_parent(dentry); + dir = d_inode(parent); /* always trust cached snapped dentries, snapdir dentry */ if (ceph_snap(dir) != CEPH_NOSNAP) { @@ -1121,13 +1111,48 @@ static int ceph_d_revalidate(struct dentry *dentry, unsigned int flags) valid = 1; } + if (!valid) { + struct ceph_mds_client *mdsc = + ceph_sb_to_client(dir->i_sb)->mdsc; + struct ceph_mds_request *req; + int op, mask, err; + + op = ceph_snap(dir) == CEPH_SNAPDIR ? + CEPH_MDS_OP_LOOKUPSNAP : CEPH_MDS_OP_LOOKUP; + req = ceph_mdsc_create_request(mdsc, op, USE_ANY_MDS); + if (!IS_ERR(req)) { + req->r_dentry = dget(dentry); + req->r_num_caps = 2; + + mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED; + if (ceph_security_xattr_wanted(dir)) + mask |= CEPH_CAP_XATTR_SHARED; + req->r_args.getattr.mask = mask; + + req->r_locked_dir = dir; + err = ceph_mdsc_do_request(mdsc, NULL, req); + if (err == 0 || err == -ENOENT) { + if (dentry == req->r_dentry) { + valid = !d_unhashed(dentry); + } else { + d_invalidate(req->r_dentry); + err = -EAGAIN; + } + } + ceph_mdsc_put_request(req); + dout("d_revalidate %p lookup result=%d\n", + dentry, err); + } + } + dout("d_revalidate %p %s\n", dentry, valid ? "valid" : "invalid"); if (valid) { ceph_dentry_lru_touch(dentry); } else { ceph_dir_clear_complete(dir); } - iput(dir); + + dput(parent); return valid; } diff --git a/fs/ceph/export.c b/fs/ceph/export.c index 3b3172357326..6e72c98162d5 100644 --- a/fs/ceph/export.c +++ b/fs/ceph/export.c @@ -71,12 +71,18 @@ static struct dentry *__fh_to_dentry(struct super_block *sb, u64 ino) inode = ceph_find_inode(sb, vino); if (!inode) { struct ceph_mds_request *req; + int mask; req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO, USE_ANY_MDS); if (IS_ERR(req)) return ERR_CAST(req); + mask = CEPH_STAT_CAP_INODE; + if (ceph_security_xattr_wanted(d_inode(sb->s_root))) + mask |= CEPH_CAP_XATTR_SHARED; + req->r_args.getattr.mask = cpu_to_le32(mask); + req->r_ino1 = vino; req->r_num_caps = 1; err = ceph_mdsc_do_request(mdsc, NULL, req); @@ -128,6 +134,7 @@ static struct dentry *__get_parent(struct super_block *sb, struct ceph_mds_request *req; struct inode *inode; struct dentry *dentry; + int mask; int err; req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPPARENT, @@ -144,6 +151,12 @@ static struct dentry *__get_parent(struct super_block *sb, .snap = CEPH_NOSNAP, }; } + + mask = CEPH_STAT_CAP_INODE; + if (ceph_security_xattr_wanted(d_inode(sb->s_root))) + mask |= CEPH_CAP_XATTR_SHARED; + req->r_args.getattr.mask = cpu_to_le32(mask); + req->r_num_caps = 1; err = ceph_mdsc_do_request(mdsc, NULL, req); inode = req->r_target_inode; diff --git a/fs/ceph/file.c b/fs/ceph/file.c index eb9028e8cfc5..ef38f01c1795 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -157,7 +157,7 @@ static int ceph_init_file(struct inode *inode, struct file *file, int fmode) case S_IFDIR: dout("init_file %p %p 0%o (regular)\n", inode, file, inode->i_mode); - cf = kmem_cache_alloc(ceph_file_cachep, GFP_KERNEL | __GFP_ZERO); + cf = kmem_cache_zalloc(ceph_file_cachep, GFP_KERNEL); if (cf == NULL) { ceph_put_fmode(ceph_inode(inode), fmode); /* clean up */ return -ENOMEM; @@ -300,6 +300,7 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry, struct ceph_mds_request *req; struct dentry *dn; struct ceph_acls_info acls = {}; + int mask; int err; dout("atomic_open %p dentry %p '%pd' %s flags %d mode 0%o\n", @@ -335,6 +336,12 @@ int ceph_atomic_open(struct inode *dir, struct dentry *dentry, acls.pagelist = NULL; } } + + mask = CEPH_STAT_CAP_INODE | CEPH_CAP_AUTH_SHARED; + if (ceph_security_xattr_wanted(dir)) + mask |= CEPH_CAP_XATTR_SHARED; + req->r_args.open.mask = cpu_to_le32(mask); + req->r_locked_dir = dir; /* caller holds dir->i_mutex */ err = ceph_mdsc_do_request(mdsc, (flags & (O_CREAT|O_TRUNC)) ? dir : NULL, @@ -725,7 +732,6 @@ static void ceph_aio_retry_work(struct work_struct *work) ret = ceph_osdc_start_request(req->r_osdc, req, false); out: if (ret < 0) { - BUG_ON(ret == -EOLDSNAPC); req->r_result = ret; ceph_aio_complete_req(req, NULL); } @@ -783,7 +789,7 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, int num_pages = 0; int flags; int ret; - struct timespec mtime = CURRENT_TIME; + struct timespec mtime = current_fs_time(inode->i_sb); size_t count = iov_iter_count(iter); loff_t pos = iocb->ki_pos; bool write = iov_iter_rw(iter) == WRITE; @@ -949,7 +955,6 @@ ceph_direct_read_write(struct kiocb *iocb, struct iov_iter *iter, ret = ceph_osdc_start_request(req->r_osdc, req, false); if (ret < 0) { - BUG_ON(ret == -EOLDSNAPC); req->r_result = ret; ceph_aio_complete_req(req, NULL); } @@ -988,7 +993,7 @@ ceph_sync_write(struct kiocb *iocb, struct iov_iter *from, loff_t pos, int flags; int check_caps = 0; int ret; - struct timespec mtime = CURRENT_TIME; + struct timespec mtime = current_fs_time(inode->i_sb); size_t count = iov_iter_count(from); if (ceph_snap(file_inode(file)) != CEPH_NOSNAP) diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index e48fd8b23257..ed58b168904a 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -549,6 +549,10 @@ int ceph_fill_file_size(struct inode *inode, int issued, if (ceph_seq_cmp(truncate_seq, ci->i_truncate_seq) > 0 || (truncate_seq == ci->i_truncate_seq && size > inode->i_size)) { dout("size %lld -> %llu\n", inode->i_size, size); + if (size > 0 && S_ISDIR(inode->i_mode)) { + pr_err("fill_file_size non-zero size for directory\n"); + size = 0; + } i_size_write(inode, size); inode->i_blocks = (size + (1<<9) - 1) >> 9; ci->i_reported_size = size; @@ -1261,6 +1265,7 @@ retry_lookup: dout(" %p links to %p %llx.%llx, not %llx.%llx\n", dn, d_inode(dn), ceph_vinop(d_inode(dn)), ceph_vinop(in)); + d_invalidate(dn); have_lease = false; } @@ -1349,15 +1354,20 @@ static int fill_readdir_cache(struct inode *dir, struct dentry *dn, if (!ctl->page || pgoff != page_index(ctl->page)) { ceph_readdir_cache_release(ctl); - ctl->page = grab_cache_page(&dir->i_data, pgoff); + if (idx == 0) + ctl->page = grab_cache_page(&dir->i_data, pgoff); + else + ctl->page = find_lock_page(&dir->i_data, pgoff); if (!ctl->page) { ctl->index = -1; - return -ENOMEM; + return idx == 0 ? -ENOMEM : 0; } /* reading/filling the cache are serialized by * i_mutex, no need to use page lock */ unlock_page(ctl->page); ctl->dentries = kmap(ctl->page); + if (idx == 0) + memset(ctl->dentries, 0, PAGE_CACHE_SIZE); } if (req->r_dir_release_cnt == atomic64_read(&ci->i_release_count) && @@ -1380,7 +1390,7 @@ int ceph_readdir_prepopulate(struct ceph_mds_request *req, struct qstr dname; struct dentry *dn; struct inode *in; - int err = 0, ret, i; + int err = 0, skipped = 0, ret, i; struct inode *snapdir = NULL; struct ceph_mds_request_head *rhead = req->r_request->front.iov_base; struct ceph_dentry_info *di; @@ -1492,7 +1502,17 @@ retry_lookup: } if (d_really_is_negative(dn)) { - struct dentry *realdn = splice_dentry(dn, in); + struct dentry *realdn; + + if (ceph_security_xattr_deadlock(in)) { + dout(" skip splicing dn %p to inode %p" + " (security xattr deadlock)\n", dn, in); + iput(in); + skipped++; + goto next_item; + } + + realdn = splice_dentry(dn, in); if (IS_ERR(realdn)) { err = PTR_ERR(realdn); d_drop(dn); @@ -1509,7 +1529,7 @@ retry_lookup: req->r_session, req->r_request_started); - if (err == 0 && cache_ctl.index >= 0) { + if (err == 0 && skipped == 0 && cache_ctl.index >= 0) { ret = fill_readdir_cache(d_inode(parent), dn, &cache_ctl, req); if (ret < 0) @@ -1520,7 +1540,7 @@ next_item: dput(dn); } out: - if (err == 0) { + if (err == 0 && skipped == 0) { req->r_did_prepopulate = true; req->r_readdir_cache_idx = cache_ctl.index; } @@ -1950,7 +1970,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) if (dirtied) { inode_dirty_flags = __ceph_mark_dirty_caps(ci, dirtied, &prealloc_cf); - inode->i_ctime = CURRENT_TIME; + inode->i_ctime = current_fs_time(inode->i_sb); } release &= issued; diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index 911d64d865f1..44852c3ae531 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -1729,7 +1729,7 @@ ceph_mdsc_create_request(struct ceph_mds_client *mdsc, int op, int mode) init_completion(&req->r_safe_completion); INIT_LIST_HEAD(&req->r_unsafe_item); - req->r_stamp = CURRENT_TIME; + req->r_stamp = current_fs_time(mdsc->fsc->sb); req->r_op = op; req->r_direct_mode = mode; @@ -2540,6 +2540,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg) /* insert trace into our cache */ mutex_lock(&req->r_fill_mutex); + current->journal_info = req; err = ceph_fill_trace(mdsc->fsc->sb, req, req->r_session); if (err == 0) { if (result == 0 && (req->r_op == CEPH_MDS_OP_READDIR || @@ -2547,6 +2548,7 @@ static void handle_reply(struct ceph_mds_session *session, struct ceph_msg *msg) ceph_readdir_prepopulate(req, req->r_session); ceph_unreserve_caps(mdsc, &req->r_caps_reservation); } + current->journal_info = NULL; mutex_unlock(&req->r_fill_mutex); up_read(&mdsc->snap_rwsem); @@ -3764,7 +3766,6 @@ void ceph_mdsc_handle_map(struct ceph_mds_client *mdsc, struct ceph_msg *msg) dout("handle_map epoch %u len %d\n", epoch, (int)maplen); /* do we need it? */ - ceph_monc_got_mdsmap(&mdsc->fsc->client->monc, epoch); mutex_lock(&mdsc->mutex); if (mdsc->mdsmap && epoch <= mdsc->mdsmap->m_epoch) { dout("handle_map epoch %u <= our %u\n", @@ -3791,6 +3792,8 @@ void ceph_mdsc_handle_map(struct ceph_mds_client *mdsc, struct ceph_msg *msg) mdsc->fsc->sb->s_maxbytes = mdsc->mdsmap->m_max_file_size; __wake_requests(mdsc, &mdsc->waiting_for_map); + ceph_monc_got_map(&mdsc->fsc->client->monc, CEPH_SUB_MDSMAP, + mdsc->mdsmap->m_epoch); mutex_unlock(&mdsc->mutex); schedule_delayed(mdsc); diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 4aa7122a8d38..9caaa7ffc93f 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -296,8 +296,6 @@ static int cmpu64_rev(const void *a, const void *b) } -struct ceph_snap_context *ceph_empty_snapc; - /* * build the snap context for a given realm. */ @@ -987,17 +985,3 @@ out: up_write(&mdsc->snap_rwsem); return; } - -int __init ceph_snap_init(void) -{ - ceph_empty_snapc = ceph_create_snap_context(0, GFP_NOFS); - if (!ceph_empty_snapc) - return -ENOMEM; - ceph_empty_snapc->seq = 1; - return 0; -} - -void ceph_snap_exit(void) -{ - ceph_put_snap_context(ceph_empty_snapc); -} diff --git a/fs/ceph/super.c b/fs/ceph/super.c index ca4d5e8457f1..c973043deb0e 100644 --- a/fs/ceph/super.c +++ b/fs/ceph/super.c @@ -439,8 +439,8 @@ static int ceph_show_options(struct seq_file *m, struct dentry *root) if (fsopt->flags & CEPH_MOUNT_OPT_DIRSTAT) seq_puts(m, ",dirstat"); - if ((fsopt->flags & CEPH_MOUNT_OPT_RBYTES) == 0) - seq_puts(m, ",norbytes"); + if ((fsopt->flags & CEPH_MOUNT_OPT_RBYTES)) + seq_puts(m, ",rbytes"); if (fsopt->flags & CEPH_MOUNT_OPT_NOASYNCREADDIR) seq_puts(m, ",noasyncreaddir"); if ((fsopt->flags & CEPH_MOUNT_OPT_DCACHE) == 0) @@ -530,7 +530,7 @@ static struct ceph_fs_client *create_fs_client(struct ceph_mount_options *fsopt, goto fail; } fsc->client->extra_mon_dispatch = extra_mon_dispatch; - fsc->client->monc.want_mdsmap = 1; + ceph_monc_want_map(&fsc->client->monc, CEPH_SUB_MDSMAP, 0, true); fsc->mount_options = fsopt; @@ -793,22 +793,20 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc, struct dentry *root; int first = 0; /* first vfsmount for this super_block */ - dout("mount start\n"); + dout("mount start %p\n", fsc); mutex_lock(&fsc->client->mount_mutex); - err = __ceph_open_session(fsc->client, started); - if (err < 0) - goto out; + if (!fsc->sb->s_root) { + err = __ceph_open_session(fsc->client, started); + if (err < 0) + goto out; - dout("mount opening root\n"); - root = open_root_dentry(fsc, "", started); - if (IS_ERR(root)) { - err = PTR_ERR(root); - goto out; - } - if (fsc->sb->s_root) { - dput(root); - } else { + dout("mount opening root\n"); + root = open_root_dentry(fsc, "", started); + if (IS_ERR(root)) { + err = PTR_ERR(root); + goto out; + } fsc->sb->s_root = root; first = 1; @@ -818,6 +816,7 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc, } if (path[0] == 0) { + root = fsc->sb->s_root; dget(root); } else { dout("mount opening base mountpoint\n"); @@ -833,16 +832,14 @@ static struct dentry *ceph_real_mount(struct ceph_fs_client *fsc, mutex_unlock(&fsc->client->mount_mutex); return root; -out: - mutex_unlock(&fsc->client->mount_mutex); - return ERR_PTR(err); - fail: if (first) { dput(fsc->sb->s_root); fsc->sb->s_root = NULL; } - goto out; +out: + mutex_unlock(&fsc->client->mount_mutex); + return ERR_PTR(err); } static int ceph_set_super(struct super_block *s, void *data) @@ -1042,19 +1039,14 @@ static int __init init_ceph(void) ceph_flock_init(); ceph_xattr_init(); - ret = ceph_snap_init(); - if (ret) - goto out_xattr; ret = register_filesystem(&ceph_fs_type); if (ret) - goto out_snap; + goto out_xattr; pr_info("loaded (mds proto %d)\n", CEPH_MDSC_PROTOCOL); return 0; -out_snap: - ceph_snap_exit(); out_xattr: ceph_xattr_exit(); destroy_caches(); @@ -1066,7 +1058,6 @@ static void __exit exit_ceph(void) { dout("exit_ceph\n"); unregister_filesystem(&ceph_fs_type); - ceph_snap_exit(); ceph_xattr_exit(); destroy_caches(); } diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 9c458eb52245..e705c4d612d7 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -37,8 +37,7 @@ #define CEPH_MOUNT_OPT_FSCACHE (1<<10) /* use fscache */ #define CEPH_MOUNT_OPT_NOPOOLPERM (1<<11) /* no pool permission check */ -#define CEPH_MOUNT_OPT_DEFAULT (CEPH_MOUNT_OPT_RBYTES | \ - CEPH_MOUNT_OPT_DCACHE) +#define CEPH_MOUNT_OPT_DEFAULT CEPH_MOUNT_OPT_DCACHE #define ceph_set_mount_opt(fsc, opt) \ (fsc)->mount_options->flags |= CEPH_MOUNT_OPT_##opt; @@ -469,7 +468,7 @@ static inline struct inode *ceph_find_inode(struct super_block *sb, #define CEPH_I_POOL_PERM (1 << 4) /* pool rd/wr bits are valid */ #define CEPH_I_POOL_RD (1 << 5) /* can read from pool */ #define CEPH_I_POOL_WR (1 << 6) /* can write to pool */ - +#define CEPH_I_SEC_INITED (1 << 7) /* security initialized */ static inline void __ceph_dir_set_complete(struct ceph_inode_info *ci, long long release_count, @@ -721,7 +720,6 @@ static inline int default_congestion_kb(void) /* snap.c */ -extern struct ceph_snap_context *ceph_empty_snapc; struct ceph_snap_realm *ceph_lookup_snap_realm(struct ceph_mds_client *mdsc, u64 ino); extern void ceph_get_snap_realm(struct ceph_mds_client *mdsc, @@ -738,8 +736,6 @@ extern void ceph_queue_cap_snap(struct ceph_inode_info *ci); extern int __ceph_finish_cap_snap(struct ceph_inode_info *ci, struct ceph_cap_snap *capsnap); extern void ceph_cleanup_empty_realms(struct ceph_mds_client *mdsc); -extern int ceph_snap_init(void); -extern void ceph_snap_exit(void); /* * a cap_snap is "pending" if it is still awaiting an in-progress @@ -808,6 +804,20 @@ extern void __init ceph_xattr_init(void); extern void ceph_xattr_exit(void); extern const struct xattr_handler *ceph_xattr_handlers[]; +#ifdef CONFIG_SECURITY +extern bool ceph_security_xattr_deadlock(struct inode *in); +extern bool ceph_security_xattr_wanted(struct inode *in); +#else +static inline bool ceph_security_xattr_deadlock(struct inode *in) +{ + return false; +} +static inline bool ceph_security_xattr_wanted(struct inode *in) +{ + return false; +} +#endif + /* acl.c */ struct ceph_acls_info { void *default_acl; @@ -947,7 +957,6 @@ extern void ceph_dentry_lru_touch(struct dentry *dn); extern void ceph_dentry_lru_del(struct dentry *dn); extern void ceph_invalidate_dentry_lease(struct dentry *dentry); extern unsigned ceph_dentry_hash(struct inode *dir, struct dentry *dn); -extern struct inode *ceph_get_dentry_parent_inode(struct dentry *dentry); extern void ceph_readdir_cache_release(struct ceph_readdir_cache_control *ctl); /* diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 819163d8313b..9410abdef3ce 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -714,31 +714,62 @@ void __ceph_build_xattrs_blob(struct ceph_inode_info *ci) } } +static inline int __get_request_mask(struct inode *in) { + struct ceph_mds_request *req = current->journal_info; + int mask = 0; + if (req && req->r_target_inode == in) { + if (req->r_op == CEPH_MDS_OP_LOOKUP || + req->r_op == CEPH_MDS_OP_LOOKUPINO || + req->r_op == CEPH_MDS_OP_LOOKUPPARENT || + req->r_op == CEPH_MDS_OP_GETATTR) { + mask = le32_to_cpu(req->r_args.getattr.mask); + } else if (req->r_op == CEPH_MDS_OP_OPEN || + req->r_op == CEPH_MDS_OP_CREATE) { + mask = le32_to_cpu(req->r_args.open.mask); + } + } + return mask; +} + ssize_t __ceph_getxattr(struct inode *inode, const char *name, void *value, size_t size) { struct ceph_inode_info *ci = ceph_inode(inode); - int err; struct ceph_inode_xattr *xattr; struct ceph_vxattr *vxattr = NULL; + int req_mask; + int err; if (!ceph_is_valid_xattr(name)) return -ENODATA; /* let's see if a virtual xattr was requested */ vxattr = ceph_match_vxattr(inode, name); - if (vxattr && !(vxattr->exists_cb && !vxattr->exists_cb(ci))) { - err = vxattr->getxattr_cb(ci, value, size); + if (vxattr) { + err = -ENODATA; + if (!(vxattr->exists_cb && !vxattr->exists_cb(ci))) + err = vxattr->getxattr_cb(ci, value, size); return err; } + req_mask = __get_request_mask(inode); + spin_lock(&ci->i_ceph_lock); dout("getxattr %p ver=%lld index_ver=%lld\n", inode, ci->i_xattrs.version, ci->i_xattrs.index_version); if (ci->i_xattrs.version == 0 || - !__ceph_caps_issued_mask(ci, CEPH_CAP_XATTR_SHARED, 1)) { + !((req_mask & CEPH_CAP_XATTR_SHARED) || + __ceph_caps_issued_mask(ci, CEPH_CAP_XATTR_SHARED, 1))) { spin_unlock(&ci->i_ceph_lock); + + /* security module gets xattr while filling trace */ + if (current->journal_info != NULL) { + pr_warn_ratelimited("sync getxattr %p " + "during filling trace\n", inode); + return -EBUSY; + } + /* get xattrs from mds (if we don't already have them) */ err = ceph_do_getattr(inode, CEPH_STAT_CAP_XATTR, true); if (err) @@ -765,6 +796,9 @@ ssize_t __ceph_getxattr(struct inode *inode, const char *name, void *value, memcpy(value, xattr->val, xattr->val_len); + if (current->journal_info != NULL && + !strncmp(name, XATTR_SECURITY_PREFIX, XATTR_SECURITY_PREFIX_LEN)) + ci->i_ceph_flags |= CEPH_I_SEC_INITED; out: spin_unlock(&ci->i_ceph_lock); return err; @@ -999,7 +1033,7 @@ retry: dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL, &prealloc_cf); ci->i_xattrs.dirty = true; - inode->i_ctime = CURRENT_TIME; + inode->i_ctime = current_fs_time(inode->i_sb); } spin_unlock(&ci->i_ceph_lock); @@ -1015,7 +1049,15 @@ do_sync: do_sync_unlocked: if (lock_snap_rwsem) up_read(&mdsc->snap_rwsem); - err = ceph_sync_setxattr(dentry, name, value, size, flags); + + /* security module set xattr while filling trace */ + if (current->journal_info != NULL) { + pr_warn_ratelimited("sync setxattr %p " + "during filling trace\n", inode); + err = -EBUSY; + } else { + err = ceph_sync_setxattr(dentry, name, value, size, flags); + } out: ceph_free_cap_flush(prealloc_cf); kfree(newname); @@ -1136,7 +1178,7 @@ retry: dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL, &prealloc_cf); ci->i_xattrs.dirty = true; - inode->i_ctime = CURRENT_TIME; + inode->i_ctime = current_fs_time(inode->i_sb); spin_unlock(&ci->i_ceph_lock); if (lock_snap_rwsem) up_read(&mdsc->snap_rwsem); @@ -1164,3 +1206,25 @@ int ceph_removexattr(struct dentry *dentry, const char *name) return __ceph_removexattr(dentry, name); } + +#ifdef CONFIG_SECURITY +bool ceph_security_xattr_wanted(struct inode *in) +{ + return in->i_security != NULL; +} + +bool ceph_security_xattr_deadlock(struct inode *in) +{ + struct ceph_inode_info *ci; + bool ret; + if (in->i_security == NULL) + return false; + ci = ceph_inode(in); + spin_lock(&ci->i_ceph_lock); + ret = !(ci->i_ceph_flags & CEPH_I_SEC_INITED) && + !(ci->i_xattrs.version > 0 && + __ceph_caps_issued_mask(ci, CEPH_CAP_XATTR_SHARED, 0)); + spin_unlock(&ci->i_ceph_lock); + return ret; +} +#endif diff --git a/fs/crypto/crypto.c b/fs/crypto/crypto.c index aed9cccca505..06cd1a22240b 100644 --- a/fs/crypto/crypto.c +++ b/fs/crypto/crypto.c @@ -170,7 +170,7 @@ static int do_page_crypto(struct inode *inode, fscrypt_complete, &ecr); BUILD_BUG_ON(FS_XTS_TWEAK_SIZE < sizeof(index)); - memcpy(xts_tweak, &inode->i_ino, sizeof(index)); + memcpy(xts_tweak, &index, sizeof(index)); memset(&xts_tweak[sizeof(index)], 0, FS_XTS_TWEAK_SIZE - sizeof(index)); diff --git a/fs/dlm/config.c b/fs/dlm/config.c index 519112168a9e..1669f6291c95 100644 --- a/fs/dlm/config.c +++ b/fs/dlm/config.c @@ -343,13 +343,12 @@ static struct config_group *make_cluster(struct config_group *g, struct dlm_cluster *cl = NULL; struct dlm_spaces *sps = NULL; struct dlm_comms *cms = NULL; - void *gps = NULL; cl = kzalloc(sizeof(struct dlm_cluster), GFP_NOFS); sps = kzalloc(sizeof(struct dlm_spaces), GFP_NOFS); cms = kzalloc(sizeof(struct dlm_comms), GFP_NOFS); - if (!cl || !gps || !sps || !cms) + if (!cl || !sps || !cms) goto fail; config_group_init_type_name(&cl->group, name, &cluster_type); diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 5c46ed9f3e14..fee81e8768c9 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -281,13 +281,15 @@ locked_inode_to_wb_and_lock_list(struct inode *inode) wb_get(wb); spin_unlock(&inode->i_lock); spin_lock(&wb->list_lock); - wb_put(wb); /* not gonna deref it anymore */ /* i_wb may have changed inbetween, can't use inode_to_wb() */ - if (likely(wb == inode->i_wb)) - return wb; /* @inode already has ref */ + if (likely(wb == inode->i_wb)) { + wb_put(wb); /* @inode already has ref */ + return wb; + } spin_unlock(&wb->list_lock); + wb_put(wb); cpu_relax(); spin_lock(&inode->i_lock); } @@ -1337,10 +1339,10 @@ __writeback_single_inode(struct inode *inode, struct writeback_control *wbc) * we go e.g. from filesystem. Flusher thread uses __writeback_single_inode() * and does more profound writeback list handling in writeback_sb_inodes(). */ -static int -writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, - struct writeback_control *wbc) +static int writeback_single_inode(struct inode *inode, + struct writeback_control *wbc) { + struct bdi_writeback *wb; int ret = 0; spin_lock(&inode->i_lock); @@ -1378,7 +1380,8 @@ writeback_single_inode(struct inode *inode, struct bdi_writeback *wb, ret = __writeback_single_inode(inode, wbc); wbc_detach_inode(wbc); - spin_lock(&wb->list_lock); + + wb = inode_to_wb_and_lock_list(inode); spin_lock(&inode->i_lock); /* * If inode is clean, remove it from writeback lists. Otherwise don't @@ -1453,6 +1456,7 @@ static long writeback_sb_inodes(struct super_block *sb, while (!list_empty(&wb->b_io)) { struct inode *inode = wb_inode(wb->b_io.prev); + struct bdi_writeback *tmp_wb; if (inode->i_sb != sb) { if (work->sb) { @@ -1543,15 +1547,23 @@ static long writeback_sb_inodes(struct super_block *sb, cond_resched(); } - - spin_lock(&wb->list_lock); + /* + * Requeue @inode if still dirty. Be careful as @inode may + * have been switched to another wb in the meantime. + */ + tmp_wb = inode_to_wb_and_lock_list(inode); spin_lock(&inode->i_lock); if (!(inode->i_state & I_DIRTY_ALL)) wrote++; - requeue_inode(inode, wb, &wbc); + requeue_inode(inode, tmp_wb, &wbc); inode_sync_complete(inode); spin_unlock(&inode->i_lock); + if (unlikely(tmp_wb != wb)) { + spin_unlock(&tmp_wb->list_lock); + spin_lock(&wb->list_lock); + } + /* * bail out to wb_writeback() often enough to check * background threshold and other termination conditions. @@ -2338,7 +2350,6 @@ EXPORT_SYMBOL(sync_inodes_sb); */ int write_inode_now(struct inode *inode, int sync) { - struct bdi_writeback *wb = &inode_to_bdi(inode)->wb; struct writeback_control wbc = { .nr_to_write = LONG_MAX, .sync_mode = sync ? WB_SYNC_ALL : WB_SYNC_NONE, @@ -2350,7 +2361,7 @@ int write_inode_now(struct inode *inode, int sync) wbc.nr_to_write = 0; might_sleep(); - return writeback_single_inode(inode, wb, &wbc); + return writeback_single_inode(inode, &wbc); } EXPORT_SYMBOL(write_inode_now); @@ -2367,7 +2378,7 @@ EXPORT_SYMBOL(write_inode_now); */ int sync_inode(struct inode *inode, struct writeback_control *wbc) { - return writeback_single_inode(inode, &inode_to_bdi(inode)->wb, wbc); + return writeback_single_inode(inode, wbc); } EXPORT_SYMBOL(sync_inode); diff --git a/fs/jffs2/gc.c b/fs/jffs2/gc.c index 95d5880a63ee..7e553f286775 100644 --- a/fs/jffs2/gc.c +++ b/fs/jffs2/gc.c @@ -134,37 +134,59 @@ int jffs2_garbage_collect_pass(struct jffs2_sb_info *c) if (mutex_lock_interruptible(&c->alloc_sem)) return -EINTR; + for (;;) { + /* We can't start doing GC until we've finished checking + the node CRCs etc. */ + int bucket, want_ino; + spin_lock(&c->erase_completion_lock); if (!c->unchecked_size) break; - - /* We can't start doing GC yet. We haven't finished checking - the node CRCs etc. Do it now. */ - - /* checked_ino is protected by the alloc_sem */ - if (c->checked_ino > c->highest_ino && xattr) { - pr_crit("Checked all inodes but still 0x%x bytes of unchecked space?\n", - c->unchecked_size); - jffs2_dbg_dump_block_lists_nolock(c); - spin_unlock(&c->erase_completion_lock); - mutex_unlock(&c->alloc_sem); - return -ENOSPC; - } - spin_unlock(&c->erase_completion_lock); if (!xattr) xattr = jffs2_verify_xattr(c); spin_lock(&c->inocache_lock); + /* Instead of doing the inodes in numeric order, doing a lookup + * in the hash for each possible number, just walk the hash + * buckets of *existing* inodes. This means that we process + * them out-of-order, but it can be a lot faster if there's + * a sparse inode# space. Which there often is. */ + want_ino = c->check_ino; + for (bucket = c->check_ino % c->inocache_hashsize ; bucket < c->inocache_hashsize; bucket++) { + for (ic = c->inocache_list[bucket]; ic; ic = ic->next) { + if (ic->ino < want_ino) + continue; + + if (ic->state != INO_STATE_CHECKEDABSENT && + ic->state != INO_STATE_PRESENT) + goto got_next; /* with inocache_lock held */ + + jffs2_dbg(1, "Skipping ino #%u already checked\n", + ic->ino); + } + want_ino = 0; + } - ic = jffs2_get_ino_cache(c, c->checked_ino++); + /* Point c->check_ino past the end of the last bucket. */ + c->check_ino = ((c->highest_ino + c->inocache_hashsize + 1) & + ~c->inocache_hashsize) - 1; - if (!ic) { - spin_unlock(&c->inocache_lock); - continue; - } + spin_unlock(&c->inocache_lock); + + pr_crit("Checked all inodes but still 0x%x bytes of unchecked space?\n", + c->unchecked_size); + jffs2_dbg_dump_block_lists_nolock(c); + mutex_unlock(&c->alloc_sem); + return -ENOSPC; + + got_next: + /* For next time round the loop, we want c->checked_ino to indicate + * the *next* one we want to check. And since we're walking the + * buckets rather than doing it sequentially, it's: */ + c->check_ino = ic->ino + c->inocache_hashsize; if (!ic->pino_nlink) { jffs2_dbg(1, "Skipping check of ino #%d with nlink/pino zero\n", @@ -176,8 +198,6 @@ int jffs2_garbage_collect_pass(struct jffs2_sb_info *c) switch(ic->state) { case INO_STATE_CHECKEDABSENT: case INO_STATE_PRESENT: - jffs2_dbg(1, "Skipping ino #%u already checked\n", - ic->ino); spin_unlock(&c->inocache_lock); continue; @@ -196,7 +216,7 @@ int jffs2_garbage_collect_pass(struct jffs2_sb_info *c) ic->ino); /* We need to come back again for the _same_ inode. We've made no progress in this case, but that should be OK */ - c->checked_ino--; + c->check_ino = ic->ino; mutex_unlock(&c->alloc_sem); sleep_on_spinunlock(&c->inocache_wq, &c->inocache_lock); diff --git a/fs/jffs2/jffs2_fs_sb.h b/fs/jffs2/jffs2_fs_sb.h index 046fee8b6e9b..778275f48a87 100644 --- a/fs/jffs2/jffs2_fs_sb.h +++ b/fs/jffs2/jffs2_fs_sb.h @@ -49,7 +49,7 @@ struct jffs2_sb_info { struct mtd_info *mtd; uint32_t highest_ino; - uint32_t checked_ino; + uint32_t check_ino; /* *NEXT* inode to be checked */ unsigned int flags; diff --git a/fs/jffs2/nodemgmt.c b/fs/jffs2/nodemgmt.c index b6bd4affd9ad..cda0774c2c9c 100644 --- a/fs/jffs2/nodemgmt.c +++ b/fs/jffs2/nodemgmt.c @@ -846,8 +846,8 @@ int jffs2_thread_should_wake(struct jffs2_sb_info *c) return 1; if (c->unchecked_size) { - jffs2_dbg(1, "jffs2_thread_should_wake(): unchecked_size %d, checked_ino #%d\n", - c->unchecked_size, c->checked_ino); + jffs2_dbg(1, "jffs2_thread_should_wake(): unchecked_size %d, check_ino #%d\n", + c->unchecked_size, c->check_ino); return 1; } diff --git a/fs/jffs2/wbuf.c b/fs/jffs2/wbuf.c index 5a3da3f52908..b25d28a21212 100644 --- a/fs/jffs2/wbuf.c +++ b/fs/jffs2/wbuf.c @@ -1183,22 +1183,20 @@ void jffs2_dirty_trigger(struct jffs2_sb_info *c) int jffs2_nand_flash_setup(struct jffs2_sb_info *c) { - struct nand_ecclayout *oinfo = c->mtd->ecclayout; - if (!c->mtd->oobsize) return 0; /* Cleanmarker is out-of-band, so inline size zero */ c->cleanmarker_size = 0; - if (!oinfo || oinfo->oobavail == 0) { + if (c->mtd->oobavail == 0) { pr_err("inconsistent device description\n"); return -EINVAL; } jffs2_dbg(1, "using OOB on NAND\n"); - c->oobavail = oinfo->oobavail; + c->oobavail = c->mtd->oobavail; /* Initialise write buffer */ init_rwsem(&c->wbuf_sem); diff --git a/fs/namei.c b/fs/namei.c index 794f81dce766..1d9ca2d5dff6 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1740,15 +1740,17 @@ static int walk_component(struct nameidata *nd, int flags) nd->flags); if (IS_ERR(path.dentry)) return PTR_ERR(path.dentry); - if (unlikely(d_is_negative(path.dentry))) { - dput(path.dentry); - return -ENOENT; - } + path.mnt = nd->path.mnt; err = follow_managed(&path, nd); if (unlikely(err < 0)) return err; + if (unlikely(d_is_negative(path.dentry))) { + path_to_nameidata(&path, nd); + return -ENOENT; + } + seq = 0; /* we are already out of RCU mode */ inode = d_backing_inode(path.dentry); } diff --git a/fs/nfs/blocklayout/blocklayout.c b/fs/nfs/blocklayout/blocklayout.c index 8bc870e4c467..02e4d87d2ed3 100644 --- a/fs/nfs/blocklayout/blocklayout.c +++ b/fs/nfs/blocklayout/blocklayout.c @@ -446,8 +446,8 @@ static void bl_free_layout_hdr(struct pnfs_layout_hdr *lo) kfree(bl); } -static struct pnfs_layout_hdr *bl_alloc_layout_hdr(struct inode *inode, - gfp_t gfp_flags) +static struct pnfs_layout_hdr *__bl_alloc_layout_hdr(struct inode *inode, + gfp_t gfp_flags, bool is_scsi_layout) { struct pnfs_block_layout *bl; @@ -460,9 +460,22 @@ static struct pnfs_layout_hdr *bl_alloc_layout_hdr(struct inode *inode, bl->bl_ext_ro = RB_ROOT; spin_lock_init(&bl->bl_ext_lock); + bl->bl_scsi_layout = is_scsi_layout; return &bl->bl_layout; } +static struct pnfs_layout_hdr *bl_alloc_layout_hdr(struct inode *inode, + gfp_t gfp_flags) +{ + return __bl_alloc_layout_hdr(inode, gfp_flags, false); +} + +static struct pnfs_layout_hdr *sl_alloc_layout_hdr(struct inode *inode, + gfp_t gfp_flags) +{ + return __bl_alloc_layout_hdr(inode, gfp_flags, true); +} + static void bl_free_lseg(struct pnfs_layout_segment *lseg) { dprintk("%s enter\n", __func__); @@ -889,22 +902,53 @@ static struct pnfs_layoutdriver_type blocklayout_type = { .sync = pnfs_generic_sync, }; +static struct pnfs_layoutdriver_type scsilayout_type = { + .id = LAYOUT_SCSI, + .name = "LAYOUT_SCSI", + .owner = THIS_MODULE, + .flags = PNFS_LAYOUTRET_ON_SETATTR | + PNFS_READ_WHOLE_PAGE, + .read_pagelist = bl_read_pagelist, + .write_pagelist = bl_write_pagelist, + .alloc_layout_hdr = sl_alloc_layout_hdr, + .free_layout_hdr = bl_free_layout_hdr, + .alloc_lseg = bl_alloc_lseg, + .free_lseg = bl_free_lseg, + .return_range = bl_return_range, + .prepare_layoutcommit = bl_prepare_layoutcommit, + .cleanup_layoutcommit = bl_cleanup_layoutcommit, + .set_layoutdriver = bl_set_layoutdriver, + .alloc_deviceid_node = bl_alloc_deviceid_node, + .free_deviceid_node = bl_free_deviceid_node, + .pg_read_ops = &bl_pg_read_ops, + .pg_write_ops = &bl_pg_write_ops, + .sync = pnfs_generic_sync, +}; + + static int __init nfs4blocklayout_init(void) { int ret; dprintk("%s: NFSv4 Block Layout Driver Registering...\n", __func__); - ret = pnfs_register_layoutdriver(&blocklayout_type); + ret = bl_init_pipefs(); if (ret) goto out; - ret = bl_init_pipefs(); + + ret = pnfs_register_layoutdriver(&blocklayout_type); if (ret) - goto out_unregister; + goto out_cleanup_pipe; + + ret = pnfs_register_layoutdriver(&scsilayout_type); + if (ret) + goto out_unregister_block; return 0; -out_unregister: +out_unregister_block: pnfs_unregister_layoutdriver(&blocklayout_type); +out_cleanup_pipe: + bl_cleanup_pipefs(); out: return ret; } @@ -914,8 +958,9 @@ static void __exit nfs4blocklayout_exit(void) dprintk("%s: NFSv4 Block Layout Driver Unregistering...\n", __func__); - bl_cleanup_pipefs(); + pnfs_unregister_layoutdriver(&scsilayout_type); pnfs_unregister_layoutdriver(&blocklayout_type); + bl_cleanup_pipefs(); } MODULE_ALIAS("nfs-layouttype4-3"); diff --git a/fs/nfs/blocklayout/blocklayout.h b/fs/nfs/blocklayout/blocklayout.h index c556640dcf3b..bc21205309e0 100644 --- a/fs/nfs/blocklayout/blocklayout.h +++ b/fs/nfs/blocklayout/blocklayout.h @@ -55,7 +55,6 @@ struct pnfs_block_dev; */ #define PNFS_BLOCK_UUID_LEN 128 - struct pnfs_block_volume { enum pnfs_block_volume_type type; union { @@ -82,6 +81,13 @@ struct pnfs_block_volume { u32 volumes_count; u32 volumes[PNFS_BLOCK_MAX_DEVICES]; } stripe; + struct { + enum scsi_code_set code_set; + enum scsi_designator_type designator_type; + int designator_len; + u8 designator[256]; + u64 pr_key; + } scsi; }; }; @@ -106,6 +112,9 @@ struct pnfs_block_dev { struct block_device *bdev; u64 disk_offset; + u64 pr_key; + bool pr_registered; + bool (*map)(struct pnfs_block_dev *dev, u64 offset, struct pnfs_block_dev_map *map); }; @@ -131,6 +140,7 @@ struct pnfs_block_layout { struct rb_root bl_ext_rw; struct rb_root bl_ext_ro; spinlock_t bl_ext_lock; /* Protects list manipulation */ + bool bl_scsi_layout; }; static inline struct pnfs_block_layout * @@ -182,6 +192,6 @@ void ext_tree_mark_committed(struct nfs4_layoutcommit_args *arg, int status); dev_t bl_resolve_deviceid(struct nfs_server *server, struct pnfs_block_volume *b, gfp_t gfp_mask); int __init bl_init_pipefs(void); -void __exit bl_cleanup_pipefs(void); +void bl_cleanup_pipefs(void); #endif /* FS_NFS_NFS4BLOCKLAYOUT_H */ diff --git a/fs/nfs/blocklayout/dev.c b/fs/nfs/blocklayout/dev.c index a861bbdfe577..e5b89675263e 100644 --- a/fs/nfs/blocklayout/dev.c +++ b/fs/nfs/blocklayout/dev.c @@ -1,11 +1,12 @@ /* - * Copyright (c) 2014 Christoph Hellwig. + * Copyright (c) 2014-2016 Christoph Hellwig. */ #include <linux/sunrpc/svc.h> #include <linux/blkdev.h> #include <linux/nfs4.h> #include <linux/nfs_fs.h> #include <linux/nfs_xdr.h> +#include <linux/pr.h> #include "blocklayout.h" @@ -21,6 +22,17 @@ bl_free_device(struct pnfs_block_dev *dev) bl_free_device(&dev->children[i]); kfree(dev->children); } else { + if (dev->pr_registered) { + const struct pr_ops *ops = + dev->bdev->bd_disk->fops->pr_ops; + int error; + + error = ops->pr_register(dev->bdev, dev->pr_key, 0, + false); + if (error) + pr_err("failed to unregister PR key.\n"); + } + if (dev->bdev) blkdev_put(dev->bdev, FMODE_READ | FMODE_WRITE); } @@ -113,6 +125,24 @@ nfs4_block_decode_volume(struct xdr_stream *xdr, struct pnfs_block_volume *b) for (i = 0; i < b->stripe.volumes_count; i++) b->stripe.volumes[i] = be32_to_cpup(p++); break; + case PNFS_BLOCK_VOLUME_SCSI: + p = xdr_inline_decode(xdr, 4 + 4 + 4); + if (!p) + return -EIO; + b->scsi.code_set = be32_to_cpup(p++); + b->scsi.designator_type = be32_to_cpup(p++); + b->scsi.designator_len = be32_to_cpup(p++); + p = xdr_inline_decode(xdr, b->scsi.designator_len); + if (!p) + return -EIO; + if (b->scsi.designator_len > 256) + return -EIO; + memcpy(&b->scsi.designator, p, b->scsi.designator_len); + p = xdr_inline_decode(xdr, 8); + if (!p) + return -EIO; + p = xdr_decode_hyper(p, &b->scsi.pr_key); + break; default: dprintk("unknown volume type!\n"); return -EIO; @@ -216,6 +246,116 @@ bl_parse_simple(struct nfs_server *server, struct pnfs_block_dev *d, return 0; } +static bool +bl_validate_designator(struct pnfs_block_volume *v) +{ + switch (v->scsi.designator_type) { + case PS_DESIGNATOR_EUI64: + if (v->scsi.code_set != PS_CODE_SET_BINARY) + return false; + + if (v->scsi.designator_len != 8 && + v->scsi.designator_len != 10 && + v->scsi.designator_len != 16) + return false; + + return true; + case PS_DESIGNATOR_NAA: + if (v->scsi.code_set != PS_CODE_SET_BINARY) + return false; + + if (v->scsi.designator_len != 8 && + v->scsi.designator_len != 16) + return false; + + return true; + case PS_DESIGNATOR_T10: + case PS_DESIGNATOR_NAME: + pr_err("pNFS: unsupported designator " + "(code set %d, type %d, len %d.\n", + v->scsi.code_set, + v->scsi.designator_type, + v->scsi.designator_len); + return false; + default: + pr_err("pNFS: invalid designator " + "(code set %d, type %d, len %d.\n", + v->scsi.code_set, + v->scsi.designator_type, + v->scsi.designator_len); + return false; + } +} + +static int +bl_parse_scsi(struct nfs_server *server, struct pnfs_block_dev *d, + struct pnfs_block_volume *volumes, int idx, gfp_t gfp_mask) +{ + struct pnfs_block_volume *v = &volumes[idx]; + const struct pr_ops *ops; + const char *devname; + int error; + + if (!bl_validate_designator(v)) + return -EINVAL; + + switch (v->scsi.designator_len) { + case 8: + devname = kasprintf(GFP_KERNEL, "/dev/disk/by-id/wwn-0x%8phN", + v->scsi.designator); + break; + case 12: + devname = kasprintf(GFP_KERNEL, "/dev/disk/by-id/wwn-0x%12phN", + v->scsi.designator); + break; + case 16: + devname = kasprintf(GFP_KERNEL, "/dev/disk/by-id/wwn-0x%16phN", + v->scsi.designator); + break; + default: + return -EINVAL; + } + + d->bdev = blkdev_get_by_path(devname, FMODE_READ, NULL); + if (IS_ERR(d->bdev)) { + pr_warn("pNFS: failed to open device %s (%ld)\n", + devname, PTR_ERR(d->bdev)); + kfree(devname); + return PTR_ERR(d->bdev); + } + + kfree(devname); + + d->len = i_size_read(d->bdev->bd_inode); + d->map = bl_map_simple; + d->pr_key = v->scsi.pr_key; + + pr_info("pNFS: using block device %s (reservation key 0x%llx)\n", + d->bdev->bd_disk->disk_name, d->pr_key); + + ops = d->bdev->bd_disk->fops->pr_ops; + if (!ops) { + pr_err("pNFS: block device %s does not support reservations.", + d->bdev->bd_disk->disk_name); + error = -EINVAL; + goto out_blkdev_put; + } + + error = ops->pr_register(d->bdev, 0, d->pr_key, true); + if (error) { + pr_err("pNFS: failed to register key for block device %s.", + d->bdev->bd_disk->disk_name); + goto out_blkdev_put; + } + + d->pr_registered = true; + return 0; + +out_blkdev_put: + blkdev_put(d->bdev, FMODE_READ); + return error; +} + static int bl_parse_slice(struct nfs_server *server, struct pnfs_block_dev *d, struct pnfs_block_volume *volumes, int idx, gfp_t gfp_mask) @@ -303,6 +443,8 @@ bl_parse_deviceid(struct nfs_server *server, struct pnfs_block_dev *d, return bl_parse_concat(server, d, volumes, idx, gfp_mask); case PNFS_BLOCK_VOLUME_STRIPE: return bl_parse_stripe(server, d, volumes, idx, gfp_mask); + case PNFS_BLOCK_VOLUME_SCSI: + return bl_parse_scsi(server, d, volumes, idx, gfp_mask); default: dprintk("unsupported volume type: %d\n", volumes[idx].type); return -EIO; diff --git a/fs/nfs/blocklayout/extent_tree.c b/fs/nfs/blocklayout/extent_tree.c index 35ab51c04814..720b3ff55fa9 100644 --- a/fs/nfs/blocklayout/extent_tree.c +++ b/fs/nfs/blocklayout/extent_tree.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2014 Christoph Hellwig. + * Copyright (c) 2014-2016 Christoph Hellwig. */ #include <linux/vmalloc.h> @@ -462,10 +462,12 @@ out: return err; } -static size_t ext_tree_layoutupdate_size(size_t count) +static size_t ext_tree_layoutupdate_size(struct pnfs_block_layout *bl, size_t count) { - return sizeof(__be32) /* number of entries */ + - PNFS_BLOCK_EXTENT_SIZE * count; + if (bl->bl_scsi_layout) + return sizeof(__be32) + PNFS_SCSI_RANGE_SIZE * count; + else + return sizeof(__be32) + PNFS_BLOCK_EXTENT_SIZE * count; } static void ext_tree_free_commitdata(struct nfs4_layoutcommit_args *arg, @@ -483,6 +485,23 @@ static void ext_tree_free_commitdata(struct nfs4_layoutcommit_args *arg, } } +static __be32 *encode_block_extent(struct pnfs_block_extent *be, __be32 *p) +{ + p = xdr_encode_opaque_fixed(p, be->be_device->deviceid.data, + NFS4_DEVICEID4_SIZE); + p = xdr_encode_hyper(p, be->be_f_offset << SECTOR_SHIFT); + p = xdr_encode_hyper(p, be->be_length << SECTOR_SHIFT); + p = xdr_encode_hyper(p, 0LL); + *p++ = cpu_to_be32(PNFS_BLOCK_READWRITE_DATA); + return p; +} + +static __be32 *encode_scsi_range(struct pnfs_block_extent *be, __be32 *p) +{ + p = xdr_encode_hyper(p, be->be_f_offset << SECTOR_SHIFT); + return xdr_encode_hyper(p, be->be_length << SECTOR_SHIFT); +} + static int ext_tree_encode_commit(struct pnfs_block_layout *bl, __be32 *p, size_t buffer_size, size_t *count) { @@ -496,19 +515,16 @@ static int ext_tree_encode_commit(struct pnfs_block_layout *bl, __be32 *p, continue; (*count)++; - if (ext_tree_layoutupdate_size(*count) > buffer_size) { + if (ext_tree_layoutupdate_size(bl, *count) > buffer_size) { /* keep counting.. */ ret = -ENOSPC; continue; } - p = xdr_encode_opaque_fixed(p, be->be_device->deviceid.data, - NFS4_DEVICEID4_SIZE); - p = xdr_encode_hyper(p, be->be_f_offset << SECTOR_SHIFT); - p = xdr_encode_hyper(p, be->be_length << SECTOR_SHIFT); - p = xdr_encode_hyper(p, 0LL); - *p++ = cpu_to_be32(PNFS_BLOCK_READWRITE_DATA); - + if (bl->bl_scsi_layout) + p = encode_scsi_range(be, p); + else + p = encode_block_extent(be, p); be->be_tag = EXTENT_COMMITTING; } spin_unlock(&bl->bl_ext_lock); @@ -537,7 +553,7 @@ retry: if (unlikely(ret)) { ext_tree_free_commitdata(arg, buffer_size); - buffer_size = ext_tree_layoutupdate_size(count); + buffer_size = ext_tree_layoutupdate_size(bl, count); count = 0; arg->layoutupdate_pages = @@ -556,7 +572,7 @@ retry: } *start_p = cpu_to_be32(count); - arg->layoutupdate_len = ext_tree_layoutupdate_size(count); + arg->layoutupdate_len = ext_tree_layoutupdate_size(bl, count); if (unlikely(arg->layoutupdate_pages != &arg->layoutupdate_page)) { void *p = start_p, *end = p + arg->layoutupdate_len; diff --git a/fs/nfs/blocklayout/rpc_pipefs.c b/fs/nfs/blocklayout/rpc_pipefs.c index dbe5839cdeba..9fb067a6f7e0 100644 --- a/fs/nfs/blocklayout/rpc_pipefs.c +++ b/fs/nfs/blocklayout/rpc_pipefs.c @@ -281,7 +281,7 @@ out: return ret; } -void __exit bl_cleanup_pipefs(void) +void bl_cleanup_pipefs(void) { rpc_pipefs_notifier_unregister(&nfs4blocklayout_block); unregister_pernet_subsys(&nfs4blocklayout_net_ops); diff --git a/fs/nfsd/Kconfig b/fs/nfsd/Kconfig index a0b77fc1bd39..c9f583d7bac8 100644 --- a/fs/nfsd/Kconfig +++ b/fs/nfsd/Kconfig @@ -84,12 +84,30 @@ config NFSD_V4 If unsure, say N. config NFSD_PNFS - bool "NFSv4.1 server support for Parallel NFS (pNFS)" - depends on NFSD_V4 + bool + +config NFSD_BLOCKLAYOUT + bool "NFSv4.1 server support for pNFS block layouts" + depends on NFSD_V4 && BLOCK + select NFSD_PNFS + help + This option enables support for the exporting pNFS block layouts + in the kernel's NFS server. The pNFS block layout enables NFS + clients to directly perform I/O to block devices accesible to both + the server and the clients. See RFC 5663 for more details. + + If unsure, say N. + +config NFSD_SCSILAYOUT + bool "NFSv4.1 server support for pNFS SCSI layouts" + depends on NFSD_V4 && BLOCK + select NFSD_PNFS help - This option enables support for the parallel NFS features of the - minor version 1 of the NFSv4 protocol (RFC5661) in the kernel's NFS - server. + This option enables support for the exporting pNFS SCSI layouts + in the kernel's NFS server. The pNFS SCSI layout enables NFS + clients to directly perform I/O to SCSI devices accesible to both + the server and the clients. See draft-ietf-nfsv4-scsi-layout for + more details. If unsure, say N. diff --git a/fs/nfsd/Makefile b/fs/nfsd/Makefile index 9a6028e120c6..3ae5f3c77e28 100644 --- a/fs/nfsd/Makefile +++ b/fs/nfsd/Makefile @@ -17,4 +17,6 @@ nfsd-$(CONFIG_NFSD_V3) += nfs3proc.o nfs3xdr.o nfsd-$(CONFIG_NFSD_V3_ACL) += nfs3acl.o nfsd-$(CONFIG_NFSD_V4) += nfs4proc.o nfs4xdr.o nfs4state.o nfs4idmap.o \ nfs4acl.o nfs4callback.o nfs4recover.o -nfsd-$(CONFIG_NFSD_PNFS) += nfs4layouts.o blocklayout.o blocklayoutxdr.o +nfsd-$(CONFIG_NFSD_PNFS) += nfs4layouts.o +nfsd-$(CONFIG_NFSD_BLOCKLAYOUT) += blocklayout.o blocklayoutxdr.o +nfsd-$(CONFIG_NFSD_SCSILAYOUT) += blocklayout.o blocklayoutxdr.o diff --git a/fs/nfsd/blocklayout.c b/fs/nfsd/blocklayout.c index c29d9421bd5e..e55b5242614d 100644 --- a/fs/nfsd/blocklayout.c +++ b/fs/nfsd/blocklayout.c @@ -1,11 +1,14 @@ /* - * Copyright (c) 2014 Christoph Hellwig. + * Copyright (c) 2014-2016 Christoph Hellwig. */ #include <linux/exportfs.h> #include <linux/genhd.h> #include <linux/slab.h> +#include <linux/pr.h> #include <linux/nfsd/debug.h> +#include <scsi/scsi_proto.h> +#include <scsi/scsi_common.h> #include "blocklayoutxdr.h" #include "pnfs.h" @@ -13,37 +16,6 @@ #define NFSDDBG_FACILITY NFSDDBG_PNFS -static int -nfsd4_block_get_device_info_simple(struct super_block *sb, - struct nfsd4_getdeviceinfo *gdp) -{ - struct pnfs_block_deviceaddr *dev; - struct pnfs_block_volume *b; - - dev = kzalloc(sizeof(struct pnfs_block_deviceaddr) + - sizeof(struct pnfs_block_volume), GFP_KERNEL); - if (!dev) - return -ENOMEM; - gdp->gd_device = dev; - - dev->nr_volumes = 1; - b = &dev->volumes[0]; - - b->type = PNFS_BLOCK_VOLUME_SIMPLE; - b->simple.sig_len = PNFS_BLOCK_UUID_LEN; - return sb->s_export_op->get_uuid(sb, b->simple.sig, &b->simple.sig_len, - &b->simple.offset); -} - -static __be32 -nfsd4_block_proc_getdeviceinfo(struct super_block *sb, - struct nfsd4_getdeviceinfo *gdp) -{ - if (sb->s_bdev != sb->s_bdev->bd_contains) - return nfserr_inval; - return nfserrno(nfsd4_block_get_device_info_simple(sb, gdp)); -} - static __be32 nfsd4_block_proc_layoutget(struct inode *inode, const struct svc_fh *fhp, struct nfsd4_layoutget *args) @@ -141,20 +113,13 @@ out_layoutunavailable: } static __be32 -nfsd4_block_proc_layoutcommit(struct inode *inode, - struct nfsd4_layoutcommit *lcp) +nfsd4_block_commit_blocks(struct inode *inode, struct nfsd4_layoutcommit *lcp, + struct iomap *iomaps, int nr_iomaps) { loff_t new_size = lcp->lc_last_wr + 1; struct iattr iattr = { .ia_valid = 0 }; - struct iomap *iomaps; - int nr_iomaps; int error; - nr_iomaps = nfsd4_block_decode_layoutupdate(lcp->lc_up_layout, - lcp->lc_up_len, &iomaps, 1 << inode->i_blkbits); - if (nr_iomaps < 0) - return nfserrno(nr_iomaps); - if (lcp->lc_mtime.tv_nsec == UTIME_NOW || timespec_compare(&lcp->lc_mtime, &inode->i_mtime) < 0) lcp->lc_mtime = current_fs_time(inode->i_sb); @@ -172,6 +137,54 @@ nfsd4_block_proc_layoutcommit(struct inode *inode, return nfserrno(error); } +#ifdef CONFIG_NFSD_BLOCKLAYOUT +static int +nfsd4_block_get_device_info_simple(struct super_block *sb, + struct nfsd4_getdeviceinfo *gdp) +{ + struct pnfs_block_deviceaddr *dev; + struct pnfs_block_volume *b; + + dev = kzalloc(sizeof(struct pnfs_block_deviceaddr) + + sizeof(struct pnfs_block_volume), GFP_KERNEL); + if (!dev) + return -ENOMEM; + gdp->gd_device = dev; + + dev->nr_volumes = 1; + b = &dev->volumes[0]; + + b->type = PNFS_BLOCK_VOLUME_SIMPLE; + b->simple.sig_len = PNFS_BLOCK_UUID_LEN; + return sb->s_export_op->get_uuid(sb, b->simple.sig, &b->simple.sig_len, + &b->simple.offset); +} + +static __be32 +nfsd4_block_proc_getdeviceinfo(struct super_block *sb, + struct nfs4_client *clp, + struct nfsd4_getdeviceinfo *gdp) +{ + if (sb->s_bdev != sb->s_bdev->bd_contains) + return nfserr_inval; + return nfserrno(nfsd4_block_get_device_info_simple(sb, gdp)); +} + +static __be32 +nfsd4_block_proc_layoutcommit(struct inode *inode, + struct nfsd4_layoutcommit *lcp) +{ + struct iomap *iomaps; + int nr_iomaps; + + nr_iomaps = nfsd4_block_decode_layoutupdate(lcp->lc_up_layout, + lcp->lc_up_len, &iomaps, 1 << inode->i_blkbits); + if (nr_iomaps < 0) + return nfserrno(nr_iomaps); + + return nfsd4_block_commit_blocks(inode, lcp, iomaps, nr_iomaps); +} + const struct nfsd4_layout_ops bl_layout_ops = { /* * Pretend that we send notification to the client. This is a blatant @@ -190,3 +203,206 @@ const struct nfsd4_layout_ops bl_layout_ops = { .encode_layoutget = nfsd4_block_encode_layoutget, .proc_layoutcommit = nfsd4_block_proc_layoutcommit, }; +#endif /* CONFIG_NFSD_BLOCKLAYOUT */ + +#ifdef CONFIG_NFSD_SCSILAYOUT +static int nfsd4_scsi_identify_device(struct block_device *bdev, + struct pnfs_block_volume *b) +{ + struct request_queue *q = bdev->bd_disk->queue; + struct request *rq; + size_t bufflen = 252, len, id_len; + u8 *buf, *d, type, assoc; + int error; + + buf = kzalloc(bufflen, GFP_KERNEL); + if (!buf) + return -ENOMEM; + + rq = blk_get_request(q, READ, GFP_KERNEL); + if (IS_ERR(rq)) { + error = -ENOMEM; + goto out_free_buf; + } + blk_rq_set_block_pc(rq); + + error = blk_rq_map_kern(q, rq, buf, bufflen, GFP_KERNEL); + if (error) + goto out_put_request; + + rq->cmd[0] = INQUIRY; + rq->cmd[1] = 1; + rq->cmd[2] = 0x83; + rq->cmd[3] = bufflen >> 8; + rq->cmd[4] = bufflen & 0xff; + rq->cmd_len = COMMAND_SIZE(INQUIRY); + + error = blk_execute_rq(rq->q, NULL, rq, 1); + if (error) { + pr_err("pNFS: INQUIRY 0x83 failed with: %x\n", + rq->errors); + goto out_put_request; + } + + len = (buf[2] << 8) + buf[3] + 4; + if (len > bufflen) { + pr_err("pNFS: INQUIRY 0x83 response invalid (len = %zd)\n", + len); + goto out_put_request; + } + + d = buf + 4; + for (d = buf + 4; d < buf + len; d += id_len + 4) { + id_len = d[3]; + type = d[1] & 0xf; + assoc = (d[1] >> 4) & 0x3; + + /* + * We only care about a EUI-64 and NAA designator types + * with LU association. + */ + if (assoc != 0x00) + continue; + if (type != 0x02 && type != 0x03) + continue; + if (id_len != 8 && id_len != 12 && id_len != 16) + continue; + + b->scsi.code_set = PS_CODE_SET_BINARY; + b->scsi.designator_type = type == 0x02 ? + PS_DESIGNATOR_EUI64 : PS_DESIGNATOR_NAA; + b->scsi.designator_len = id_len; + memcpy(b->scsi.designator, d + 4, id_len); + + /* + * If we found a 8 or 12 byte descriptor continue on to + * see if a 16 byte one is available. If we find a + * 16 byte descriptor we're done. + */ + if (id_len == 16) + break; + } + +out_put_request: + blk_put_request(rq); +out_free_buf: + kfree(buf); + return error; +} + +#define NFSD_MDS_PR_KEY 0x0100000000000000 + +/* + * We use the client ID as a unique key for the reservations. + * This allows us to easily fence a client when recalls fail. + */ +static u64 nfsd4_scsi_pr_key(struct nfs4_client *clp) +{ + return ((u64)clp->cl_clientid.cl_boot << 32) | clp->cl_clientid.cl_id; +} + +static int +nfsd4_block_get_device_info_scsi(struct super_block *sb, + struct nfs4_client *clp, + struct nfsd4_getdeviceinfo *gdp) +{ + struct pnfs_block_deviceaddr *dev; + struct pnfs_block_volume *b; + const struct pr_ops *ops; + int error; + + dev = kzalloc(sizeof(struct pnfs_block_deviceaddr) + + sizeof(struct pnfs_block_volume), GFP_KERNEL); + if (!dev) + return -ENOMEM; + gdp->gd_device = dev; + + dev->nr_volumes = 1; + b = &dev->volumes[0]; + + b->type = PNFS_BLOCK_VOLUME_SCSI; + b->scsi.pr_key = nfsd4_scsi_pr_key(clp); + + error = nfsd4_scsi_identify_device(sb->s_bdev, b); + if (error) + return error; + + ops = sb->s_bdev->bd_disk->fops->pr_ops; + if (!ops) { + pr_err("pNFS: device %s does not support PRs.\n", + sb->s_id); + return -EINVAL; + } + + error = ops->pr_register(sb->s_bdev, 0, NFSD_MDS_PR_KEY, true); + if (error) { + pr_err("pNFS: failed to register key for device %s.\n", + sb->s_id); + return -EINVAL; + } + + error = ops->pr_reserve(sb->s_bdev, NFSD_MDS_PR_KEY, + PR_EXCLUSIVE_ACCESS_REG_ONLY, 0); + if (error) { + pr_err("pNFS: failed to reserve device %s.\n", + sb->s_id); + return -EINVAL; + } + + return 0; +} + +static __be32 +nfsd4_scsi_proc_getdeviceinfo(struct super_block *sb, + struct nfs4_client *clp, + struct nfsd4_getdeviceinfo *gdp) +{ + if (sb->s_bdev != sb->s_bdev->bd_contains) + return nfserr_inval; + return nfserrno(nfsd4_block_get_device_info_scsi(sb, clp, gdp)); +} +static __be32 +nfsd4_scsi_proc_layoutcommit(struct inode *inode, + struct nfsd4_layoutcommit *lcp) +{ + struct iomap *iomaps; + int nr_iomaps; + + nr_iomaps = nfsd4_scsi_decode_layoutupdate(lcp->lc_up_layout, + lcp->lc_up_len, &iomaps, 1 << inode->i_blkbits); + if (nr_iomaps < 0) + return nfserrno(nr_iomaps); + + return nfsd4_block_commit_blocks(inode, lcp, iomaps, nr_iomaps); +} + +static void +nfsd4_scsi_fence_client(struct nfs4_layout_stateid *ls) +{ + struct nfs4_client *clp = ls->ls_stid.sc_client; + struct block_device *bdev = ls->ls_file->f_path.mnt->mnt_sb->s_bdev; + + bdev->bd_disk->fops->pr_ops->pr_preempt(bdev, NFSD_MDS_PR_KEY, + nfsd4_scsi_pr_key(clp), 0, true); +} + +const struct nfsd4_layout_ops scsi_layout_ops = { + /* + * Pretend that we send notification to the client. This is a blatant + * lie to force recent Linux clients to cache our device IDs. + * We rarely ever change the device ID, so the harm of leaking deviceids + * for a while isn't too bad. Unfortunately RFC5661 is a complete mess + * in this regard, but I filed errata 4119 for this a while ago, and + * hopefully the Linux client will eventually start caching deviceids + * without this again. + */ + .notify_types = + NOTIFY_DEVICEID4_DELETE | NOTIFY_DEVICEID4_CHANGE, + .proc_getdeviceinfo = nfsd4_scsi_proc_getdeviceinfo, + .encode_getdeviceinfo = nfsd4_block_encode_getdeviceinfo, + .proc_layoutget = nfsd4_block_proc_layoutget, + .encode_layoutget = nfsd4_block_encode_layoutget, + .proc_layoutcommit = nfsd4_scsi_proc_layoutcommit, + .fence_client = nfsd4_scsi_fence_client, +}; +#endif /* CONFIG_NFSD_SCSILAYOUT */ diff --git a/fs/nfsd/blocklayoutxdr.c b/fs/nfsd/blocklayoutxdr.c index 6d834dc9bbc8..6c3b316f932e 100644 --- a/fs/nfsd/blocklayoutxdr.c +++ b/fs/nfsd/blocklayoutxdr.c @@ -1,5 +1,5 @@ /* - * Copyright (c) 2014 Christoph Hellwig. + * Copyright (c) 2014-2016 Christoph Hellwig. */ #include <linux/sunrpc/svc.h> #include <linux/exportfs.h> @@ -53,6 +53,18 @@ nfsd4_block_encode_volume(struct xdr_stream *xdr, struct pnfs_block_volume *b) p = xdr_encode_hyper(p, b->simple.offset); p = xdr_encode_opaque(p, b->simple.sig, b->simple.sig_len); break; + case PNFS_BLOCK_VOLUME_SCSI: + len = 4 + 4 + 4 + 4 + b->scsi.designator_len + 8; + p = xdr_reserve_space(xdr, len); + if (!p) + return -ETOOSMALL; + + *p++ = cpu_to_be32(b->type); + *p++ = cpu_to_be32(b->scsi.code_set); + *p++ = cpu_to_be32(b->scsi.designator_type); + p = xdr_encode_opaque(p, b->scsi.designator, b->scsi.designator_len); + p = xdr_encode_hyper(p, b->scsi.pr_key); + break; default: return -ENOTSUPP; } @@ -93,18 +105,22 @@ nfsd4_block_decode_layoutupdate(__be32 *p, u32 len, struct iomap **iomapp, u32 block_size) { struct iomap *iomaps; - u32 nr_iomaps, expected, i; + u32 nr_iomaps, i; if (len < sizeof(u32)) { dprintk("%s: extent array too small: %u\n", __func__, len); return -EINVAL; } + len -= sizeof(u32); + if (len % PNFS_BLOCK_EXTENT_SIZE) { + dprintk("%s: extent array invalid: %u\n", __func__, len); + return -EINVAL; + } nr_iomaps = be32_to_cpup(p++); - expected = sizeof(__be32) + nr_iomaps * PNFS_BLOCK_EXTENT_SIZE; - if (len != expected) { + if (nr_iomaps != len / PNFS_BLOCK_EXTENT_SIZE) { dprintk("%s: extent array size mismatch: %u/%u\n", - __func__, len, expected); + __func__, len, nr_iomaps); return -EINVAL; } @@ -155,3 +171,54 @@ fail: kfree(iomaps); return -EINVAL; } + +int +nfsd4_scsi_decode_layoutupdate(__be32 *p, u32 len, struct iomap **iomapp, + u32 block_size) +{ + struct iomap *iomaps; + u32 nr_iomaps, expected, i; + + if (len < sizeof(u32)) { + dprintk("%s: extent array too small: %u\n", __func__, len); + return -EINVAL; + } + + nr_iomaps = be32_to_cpup(p++); + expected = sizeof(__be32) + nr_iomaps * PNFS_SCSI_RANGE_SIZE; + if (len != expected) { + dprintk("%s: extent array size mismatch: %u/%u\n", + __func__, len, expected); + return -EINVAL; + } + + iomaps = kcalloc(nr_iomaps, sizeof(*iomaps), GFP_KERNEL); + if (!iomaps) { + dprintk("%s: failed to allocate extent array\n", __func__); + return -ENOMEM; + } + + for (i = 0; i < nr_iomaps; i++) { + u64 val; + + p = xdr_decode_hyper(p, &val); + if (val & (block_size - 1)) { + dprintk("%s: unaligned offset 0x%llx\n", __func__, val); + goto fail; + } + iomaps[i].offset = val; + + p = xdr_decode_hyper(p, &val); + if (val & (block_size - 1)) { + dprintk("%s: unaligned length 0x%llx\n", __func__, val); + goto fail; + } + iomaps[i].length = val; + } + + *iomapp = iomaps; + return nr_iomaps; +fail: + kfree(iomaps); + return -EINVAL; +} diff --git a/fs/nfsd/blocklayoutxdr.h b/fs/nfsd/blocklayoutxdr.h index 6de925fe8499..397bc7563a49 100644 --- a/fs/nfsd/blocklayoutxdr.h +++ b/fs/nfsd/blocklayoutxdr.h @@ -15,6 +15,11 @@ struct pnfs_block_extent { enum pnfs_block_extent_state es; }; +struct pnfs_block_range { + u64 foff; + u64 len; +}; + /* * Random upper cap for the uuid length to avoid unbounded allocation. * Not actually limited by the protocol. @@ -29,6 +34,13 @@ struct pnfs_block_volume { u32 sig_len; u8 sig[PNFS_BLOCK_UUID_LEN]; } simple; + struct { + enum scsi_code_set code_set; + enum scsi_designator_type designator_type; + int designator_len; + u8 designator[256]; + u64 pr_key; + } scsi; }; }; @@ -43,5 +55,7 @@ __be32 nfsd4_block_encode_layoutget(struct xdr_stream *xdr, struct nfsd4_layoutget *lgp); int nfsd4_block_decode_layoutupdate(__be32 *p, u32 len, struct iomap **iomapp, u32 block_size); +int nfsd4_scsi_decode_layoutupdate(__be32 *p, u32 len, struct iomap **iomapp, + u32 block_size); #endif /* _NFSD_BLOCKLAYOUTXDR_H */ diff --git a/fs/nfsd/nfs3proc.c b/fs/nfsd/nfs3proc.c index 7b755b7f785c..51c3b06e8036 100644 --- a/fs/nfsd/nfs3proc.c +++ b/fs/nfsd/nfs3proc.c @@ -147,6 +147,7 @@ nfsd3_proc_read(struct svc_rqst *rqstp, struct nfsd3_readargs *argp, { __be32 nfserr; u32 max_blocksize = svc_max_payload(rqstp); + unsigned long cnt = min(argp->count, max_blocksize); dprintk("nfsd: READ(3) %s %lu bytes at %Lu\n", SVCFH_fmt(&argp->fh), @@ -157,7 +158,7 @@ nfsd3_proc_read(struct svc_rqst *rqstp, struct nfsd3_readargs *argp, * 1 (status) + 22 (post_op_attr) + 1 (count) + 1 (eof) * + 1 (xdr opaque byte count) = 26 */ - resp->count = min(argp->count, max_blocksize); + resp->count = cnt; svc_reserve_auth(rqstp, ((1 + NFS3_POST_OP_ATTR_WORDS + 3)<<2) + resp->count +4); fh_copy(&resp->fh, &argp->fh); @@ -167,8 +168,8 @@ nfsd3_proc_read(struct svc_rqst *rqstp, struct nfsd3_readargs *argp, &resp->count); if (nfserr == 0) { struct inode *inode = d_inode(resp->fh.fh_dentry); - - resp->eof = (argp->offset + resp->count) >= inode->i_size; + resp->eof = nfsd_eof_on_read(cnt, resp->count, argp->offset, + inode->i_size); } RETURN_STATUS(nfserr); diff --git a/fs/nfsd/nfs4layouts.c b/fs/nfsd/nfs4layouts.c index ce2d010d3b17..825c7bc8d789 100644 --- a/fs/nfsd/nfs4layouts.c +++ b/fs/nfsd/nfs4layouts.c @@ -1,6 +1,7 @@ /* * Copyright (c) 2014 Christoph Hellwig. */ +#include <linux/blkdev.h> #include <linux/kmod.h> #include <linux/file.h> #include <linux/jhash.h> @@ -26,7 +27,12 @@ static const struct nfsd4_callback_ops nfsd4_cb_layout_ops; static const struct lock_manager_operations nfsd4_layouts_lm_ops; const struct nfsd4_layout_ops *nfsd4_layout_ops[LAYOUT_TYPE_MAX] = { +#ifdef CONFIG_NFSD_BLOCKLAYOUT [LAYOUT_BLOCK_VOLUME] = &bl_layout_ops, +#endif +#ifdef CONFIG_NFSD_SCSILAYOUT + [LAYOUT_SCSI] = &scsi_layout_ops, +#endif }; /* pNFS device ID to export fsid mapping */ @@ -121,10 +127,24 @@ void nfsd4_setup_layout_type(struct svc_export *exp) if (!(exp->ex_flags & NFSEXP_PNFS)) return; + /* + * Check if the file system supports exporting a block-like layout. + * If the block device supports reservations prefer the SCSI layout, + * otherwise advertise the block layout. + */ +#ifdef CONFIG_NFSD_BLOCKLAYOUT if (sb->s_export_op->get_uuid && sb->s_export_op->map_blocks && sb->s_export_op->commit_blocks) exp->ex_layout_type = LAYOUT_BLOCK_VOLUME; +#endif +#ifdef CONFIG_NFSD_SCSILAYOUT + /* overwrite block layout selection if needed */ + if (sb->s_export_op->map_blocks && + sb->s_export_op->commit_blocks && + sb->s_bdev && sb->s_bdev->bd_disk->fops->pr_ops) + exp->ex_layout_type = LAYOUT_SCSI; +#endif } static void @@ -590,8 +610,6 @@ nfsd4_cb_layout_fail(struct nfs4_layout_stateid *ls) rpc_ntop((struct sockaddr *)&clp->cl_addr, addr_str, sizeof(addr_str)); - trace_layout_recall_fail(&ls->ls_stid.sc_stateid); - printk(KERN_WARNING "nfsd: client %s failed to respond to layout recall. " " Fencing..\n", addr_str); @@ -626,6 +644,7 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task) container_of(cb, struct nfs4_layout_stateid, ls_recall); struct nfsd_net *nn; ktime_t now, cutoff; + const struct nfsd4_layout_ops *ops; LIST_HEAD(reaplist); @@ -661,7 +680,13 @@ nfsd4_cb_layout_done(struct nfsd4_callback *cb, struct rpc_task *task) /* * Unknown error or non-responding client, we'll need to fence. */ - nfsd4_cb_layout_fail(ls); + trace_layout_recall_fail(&ls->ls_stid.sc_stateid); + + ops = nfsd4_layout_ops[ls->ls_layout_type]; + if (ops->fence_client) + ops->fence_client(ls); + else + nfsd4_cb_layout_fail(ls); return -1; } } diff --git a/fs/nfsd/nfs4proc.c b/fs/nfsd/nfs4proc.c index 40b912407d51..de1ff1d98bb1 100644 --- a/fs/nfsd/nfs4proc.c +++ b/fs/nfsd/nfs4proc.c @@ -1268,8 +1268,10 @@ nfsd4_getdeviceinfo(struct svc_rqst *rqstp, goto out; nfserr = nfs_ok; - if (gdp->gd_maxcount != 0) - nfserr = ops->proc_getdeviceinfo(exp->ex_path.mnt->mnt_sb, gdp); + if (gdp->gd_maxcount != 0) { + nfserr = ops->proc_getdeviceinfo(exp->ex_path.mnt->mnt_sb, + cstate->session->se_client, gdp); + } gdp->gd_notify_types &= ops->notify_types; out: diff --git a/fs/nfsd/nfs4xdr.c b/fs/nfsd/nfs4xdr.c index aa87954b4af2..9df898ba648f 100644 --- a/fs/nfsd/nfs4xdr.c +++ b/fs/nfsd/nfs4xdr.c @@ -3365,6 +3365,7 @@ static __be32 nfsd4_encode_splice_read( struct xdr_stream *xdr = &resp->xdr; struct xdr_buf *buf = xdr->buf; u32 eof; + long len; int space_left; __be32 nfserr; __be32 *p = xdr->p - 2; @@ -3373,6 +3374,7 @@ static __be32 nfsd4_encode_splice_read( if (xdr->end - xdr->p < 1) return nfserr_resource; + len = maxcount; nfserr = nfsd_splice_read(read->rd_rqstp, file, read->rd_offset, &maxcount); if (nfserr) { @@ -3385,8 +3387,8 @@ static __be32 nfsd4_encode_splice_read( return nfserr; } - eof = (read->rd_offset + maxcount >= - d_inode(read->rd_fhp->fh_dentry)->i_size); + eof = nfsd_eof_on_read(len, maxcount, read->rd_offset, + d_inode(read->rd_fhp->fh_dentry)->i_size); *(p++) = htonl(eof); *(p++) = htonl(maxcount); @@ -3456,14 +3458,15 @@ static __be32 nfsd4_encode_readv(struct nfsd4_compoundres *resp, } read->rd_vlen = v; + len = maxcount; nfserr = nfsd_readv(file, read->rd_offset, resp->rqstp->rq_vec, read->rd_vlen, &maxcount); if (nfserr) return nfserr; xdr_truncate_encode(xdr, starting_len + 8 + ((maxcount+3)&~3)); - eof = (read->rd_offset + maxcount >= - d_inode(read->rd_fhp->fh_dentry)->i_size); + eof = nfsd_eof_on_read(len, maxcount, read->rd_offset, + d_inode(read->rd_fhp->fh_dentry)->i_size); tmp = htonl(eof); write_bytes_to_xdr_buf(xdr->buf, starting_len , &tmp, 4); diff --git a/fs/nfsd/pnfs.h b/fs/nfsd/pnfs.h index d4c4453674c6..7d073b9b1553 100644 --- a/fs/nfsd/pnfs.h +++ b/fs/nfsd/pnfs.h @@ -21,6 +21,7 @@ struct nfsd4_layout_ops { u32 notify_types; __be32 (*proc_getdeviceinfo)(struct super_block *sb, + struct nfs4_client *clp, struct nfsd4_getdeviceinfo *gdevp); __be32 (*encode_getdeviceinfo)(struct xdr_stream *xdr, struct nfsd4_getdeviceinfo *gdevp); @@ -32,10 +33,17 @@ struct nfsd4_layout_ops { __be32 (*proc_layoutcommit)(struct inode *inode, struct nfsd4_layoutcommit *lcp); + + void (*fence_client)(struct nfs4_layout_stateid *ls); }; extern const struct nfsd4_layout_ops *nfsd4_layout_ops[]; +#ifdef CONFIG_NFSD_BLOCKLAYOUT extern const struct nfsd4_layout_ops bl_layout_ops; +#endif +#ifdef CONFIG_NFSD_SCSILAYOUT +extern const struct nfsd4_layout_ops scsi_layout_ops; +#endif __be32 nfsd4_preprocess_layout_stateid(struct svc_rqst *rqstp, struct nfsd4_compound_state *cstate, stateid_t *stateid, diff --git a/fs/nfsd/vfs.h b/fs/nfsd/vfs.h index c11ba316f23f..2d573ec057f8 100644 --- a/fs/nfsd/vfs.h +++ b/fs/nfsd/vfs.h @@ -139,4 +139,23 @@ static inline int nfsd_create_is_exclusive(int createmode) || createmode == NFS4_CREATE_EXCLUSIVE4_1; } +static inline bool nfsd_eof_on_read(long requested, long read, + loff_t offset, loff_t size) +{ + /* We assume a short read means eof: */ + if (requested > read) + return true; + /* + * A non-short read might also reach end of file. The spec + * still requires us to set eof in that case. + * + * Further operations may have modified the file size since + * the read, so the following check is not atomic with the read. + * We've only seen that cause a problem for a client in the case + * where the read returned a count of 0 without setting eof. + * That case was fixed by the addition of the above check. + */ + return (offset + read >= size); +} + #endif /* LINUX_NFSD_VFS_H */ diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index d002579c6f2b..70907d638b60 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -2516,21 +2516,6 @@ static int ocfs2_update_edge_lengths(handle_t *handle, struct ocfs2_extent_block *eb; u32 range; - /* - * In normal tree rotation process, we will never touch the - * tree branch above subtree_index and ocfs2_extend_rotate_transaction - * doesn't reserve the credits for them either. - * - * But we do have a special case here which will update the rightmost - * records for all the bh in the path. - * So we have to allocate extra credits and access them. - */ - ret = ocfs2_extend_trans(handle, subtree_index); - if (ret) { - mlog_errno(ret); - goto out; - } - ret = ocfs2_journal_access_path(et->et_ci, handle, path); if (ret) { mlog_errno(ret); @@ -2956,7 +2941,7 @@ static int __ocfs2_rotate_tree_left(handle_t *handle, right_path->p_node[subtree_root].bh->b_blocknr, right_path->p_tree_depth); - ret = ocfs2_extend_rotate_transaction(handle, subtree_root, + ret = ocfs2_extend_rotate_transaction(handle, 0, orig_credits, left_path); if (ret) { mlog_errno(ret); @@ -3029,21 +3014,9 @@ static int ocfs2_remove_rightmost_path(handle_t *handle, struct ocfs2_extent_block *eb; struct ocfs2_extent_list *el; - ret = ocfs2_et_sanity_check(et); if (ret) goto out; - /* - * There's two ways we handle this depending on - * whether path is the only existing one. - */ - ret = ocfs2_extend_rotate_transaction(handle, 0, - handle->h_buffer_credits, - path); - if (ret) { - mlog_errno(ret); - goto out; - } ret = ocfs2_journal_access_path(et->et_ci, handle, path); if (ret) { @@ -3641,6 +3614,14 @@ static int ocfs2_merge_rec_left(struct ocfs2_path *right_path, */ if (le16_to_cpu(right_rec->e_leaf_clusters) == 0 && le16_to_cpu(el->l_next_free_rec) == 1) { + /* extend credit for ocfs2_remove_rightmost_path */ + ret = ocfs2_extend_rotate_transaction(handle, 0, + handle->h_buffer_credits, + right_path); + if (ret) { + mlog_errno(ret); + goto out; + } ret = ocfs2_remove_rightmost_path(handle, et, right_path, @@ -3679,6 +3660,14 @@ static int ocfs2_try_to_merge_extent(handle_t *handle, BUG_ON(ctxt->c_contig_type == CONTIG_NONE); if (ctxt->c_split_covers_rec && ctxt->c_has_empty_extent) { + /* extend credit for ocfs2_remove_rightmost_path */ + ret = ocfs2_extend_rotate_transaction(handle, 0, + handle->h_buffer_credits, + path); + if (ret) { + mlog_errno(ret); + goto out; + } /* * The merge code will need to create an empty * extent to take the place of the newly @@ -3727,6 +3716,15 @@ static int ocfs2_try_to_merge_extent(handle_t *handle, */ BUG_ON(!ocfs2_is_empty_extent(&el->l_recs[0])); + /* extend credit for ocfs2_remove_rightmost_path */ + ret = ocfs2_extend_rotate_transaction(handle, 0, + handle->h_buffer_credits, + path); + if (ret) { + mlog_errno(ret); + goto out; + } + /* The merge left us with an empty extent, remove it. */ ret = ocfs2_rotate_tree_left(handle, et, path, dealloc); if (ret) { @@ -3748,6 +3746,15 @@ static int ocfs2_try_to_merge_extent(handle_t *handle, goto out; } + /* extend credit for ocfs2_remove_rightmost_path */ + ret = ocfs2_extend_rotate_transaction(handle, 0, + handle->h_buffer_credits, + path); + if (ret) { + mlog_errno(ret); + goto out; + } + ret = ocfs2_rotate_tree_left(handle, et, path, dealloc); /* * Error from this last rotate is not critical, so @@ -3783,6 +3790,16 @@ static int ocfs2_try_to_merge_extent(handle_t *handle, } if (ctxt->c_split_covers_rec) { + /* extend credit for ocfs2_remove_rightmost_path */ + ret = ocfs2_extend_rotate_transaction(handle, 0, + handle->h_buffer_credits, + path); + if (ret) { + mlog_errno(ret); + ret = 0; + goto out; + } + /* * The merge may have left an empty extent in * our leaf. Try to rotate it away. @@ -5342,6 +5359,15 @@ static int ocfs2_truncate_rec(handle_t *handle, struct ocfs2_extent_block *eb; if (ocfs2_is_empty_extent(&el->l_recs[0]) && index > 0) { + /* extend credit for ocfs2_remove_rightmost_path */ + ret = ocfs2_extend_rotate_transaction(handle, 0, + handle->h_buffer_credits, + path); + if (ret) { + mlog_errno(ret); + goto out; + } + ret = ocfs2_rotate_tree_left(handle, et, path, dealloc); if (ret) { mlog_errno(ret); @@ -5928,16 +5954,6 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, ocfs2_journal_dirty(handle, tl_bh); - /* TODO: Perhaps we can calculate the bulk of the - * credits up front rather than extending like - * this. */ - status = ocfs2_extend_trans(handle, - OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC); - if (status < 0) { - mlog_errno(status); - goto bail; - } - rec = tl->tl_recs[i]; start_blk = ocfs2_clusters_to_blocks(data_alloc_inode->i_sb, le32_to_cpu(rec.t_start)); @@ -5958,6 +5974,13 @@ static int ocfs2_replay_truncate_records(struct ocfs2_super *osb, goto bail; } } + + status = ocfs2_extend_trans(handle, + OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC); + if (status < 0) { + mlog_errno(status); + goto bail; + } i--; } @@ -6016,7 +6039,7 @@ int __ocfs2_flush_truncate_log(struct ocfs2_super *osb) goto out_mutex; } - handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_UPDATE); + handle = ocfs2_start_trans(osb, OCFS2_TRUNCATE_LOG_FLUSH_ONE_REC); if (IS_ERR(handle)) { status = PTR_ERR(handle); mlog_errno(status); @@ -6079,7 +6102,7 @@ void ocfs2_schedule_truncate_log_flush(struct ocfs2_super *osb, if (cancel) cancel_delayed_work(&osb->osb_truncate_log_wq); - queue_delayed_work(ocfs2_wq, &osb->osb_truncate_log_wq, + queue_delayed_work(osb->ocfs2_wq, &osb->osb_truncate_log_wq, OCFS2_TRUNCATE_LOG_FLUSH_INTERVAL); } } @@ -6253,7 +6276,7 @@ void ocfs2_truncate_log_shutdown(struct ocfs2_super *osb) if (tl_inode) { cancel_delayed_work(&osb->osb_truncate_log_wq); - flush_workqueue(ocfs2_wq); + flush_workqueue(osb->ocfs2_wq); status = ocfs2_flush_truncate_log(osb); if (status < 0) diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 043110e5212d..1581240a7ca0 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -499,158 +499,6 @@ bail: return status; } -/* - * TODO: Make this into a generic get_blocks function. - * - * From do_direct_io in direct-io.c: - * "So what we do is to permit the ->get_blocks function to populate - * bh.b_size with the size of IO which is permitted at this offset and - * this i_blkbits." - * - * This function is called directly from get_more_blocks in direct-io.c. - * - * called like this: dio->get_blocks(dio->inode, fs_startblk, - * fs_count, map_bh, dio->rw == WRITE); - */ -static int ocfs2_direct_IO_get_blocks(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int create) -{ - int ret; - u32 cpos = 0; - int alloc_locked = 0; - u64 p_blkno, inode_blocks, contig_blocks; - unsigned int ext_flags; - unsigned char blocksize_bits = inode->i_sb->s_blocksize_bits; - unsigned long max_blocks = bh_result->b_size >> inode->i_blkbits; - unsigned long len = bh_result->b_size; - unsigned int clusters_to_alloc = 0, contig_clusters = 0; - - cpos = ocfs2_blocks_to_clusters(inode->i_sb, iblock); - - /* This function won't even be called if the request isn't all - * nicely aligned and of the right size, so there's no need - * for us to check any of that. */ - - inode_blocks = ocfs2_blocks_for_bytes(inode->i_sb, i_size_read(inode)); - - down_read(&OCFS2_I(inode)->ip_alloc_sem); - - /* This figures out the size of the next contiguous block, and - * our logical offset */ - ret = ocfs2_extent_map_get_blocks(inode, iblock, &p_blkno, - &contig_blocks, &ext_flags); - up_read(&OCFS2_I(inode)->ip_alloc_sem); - - if (ret) { - mlog(ML_ERROR, "get_blocks() failed iblock=%llu\n", - (unsigned long long)iblock); - ret = -EIO; - goto bail; - } - - /* We should already CoW the refcounted extent in case of create. */ - BUG_ON(create && (ext_flags & OCFS2_EXT_REFCOUNTED)); - - /* allocate blocks if no p_blkno is found, and create == 1 */ - if (!p_blkno && create) { - ret = ocfs2_inode_lock(inode, NULL, 1); - if (ret < 0) { - mlog_errno(ret); - goto bail; - } - - alloc_locked = 1; - - down_write(&OCFS2_I(inode)->ip_alloc_sem); - - /* fill hole, allocate blocks can't be larger than the size - * of the hole */ - clusters_to_alloc = ocfs2_clusters_for_bytes(inode->i_sb, len); - contig_clusters = ocfs2_clusters_for_blocks(inode->i_sb, - contig_blocks); - if (clusters_to_alloc > contig_clusters) - clusters_to_alloc = contig_clusters; - - /* allocate extent and insert them into the extent tree */ - ret = ocfs2_extend_allocation(inode, cpos, - clusters_to_alloc, 0); - if (ret < 0) { - up_write(&OCFS2_I(inode)->ip_alloc_sem); - mlog_errno(ret); - goto bail; - } - - ret = ocfs2_extent_map_get_blocks(inode, iblock, &p_blkno, - &contig_blocks, &ext_flags); - if (ret < 0) { - up_write(&OCFS2_I(inode)->ip_alloc_sem); - mlog(ML_ERROR, "get_blocks() failed iblock=%llu\n", - (unsigned long long)iblock); - ret = -EIO; - goto bail; - } - set_buffer_new(bh_result); - up_write(&OCFS2_I(inode)->ip_alloc_sem); - } - - /* - * get_more_blocks() expects us to describe a hole by clearing - * the mapped bit on bh_result(). - * - * Consider an unwritten extent as a hole. - */ - if (p_blkno && !(ext_flags & OCFS2_EXT_UNWRITTEN)) - map_bh(bh_result, inode->i_sb, p_blkno); - else - clear_buffer_mapped(bh_result); - - /* make sure we don't map more than max_blocks blocks here as - that's all the kernel will handle at this point. */ - if (max_blocks < contig_blocks) - contig_blocks = max_blocks; - bh_result->b_size = contig_blocks << blocksize_bits; -bail: - if (alloc_locked) - ocfs2_inode_unlock(inode, 1); - return ret; -} - -/* - * ocfs2_dio_end_io is called by the dio core when a dio is finished. We're - * particularly interested in the aio/dio case. We use the rw_lock DLM lock - * to protect io on one node from truncation on another. - */ -static int ocfs2_dio_end_io(struct kiocb *iocb, - loff_t offset, - ssize_t bytes, - void *private) -{ - struct inode *inode = file_inode(iocb->ki_filp); - int level; - - if (bytes <= 0) - return 0; - - /* this io's submitter should not have unlocked this before we could */ - BUG_ON(!ocfs2_iocb_is_rw_locked(iocb)); - - if (ocfs2_iocb_is_unaligned_aio(iocb)) { - ocfs2_iocb_clear_unaligned_aio(iocb); - - mutex_unlock(&OCFS2_I(inode)->ip_unaligned_aio); - } - - /* Let rw unlock to be done later to protect append direct io write */ - if (offset + bytes <= i_size_read(inode)) { - ocfs2_iocb_clear_rw_locked(iocb); - - level = ocfs2_iocb_rw_locked_level(iocb); - ocfs2_rw_unlock(inode, level); - } - - return 0; -} - static int ocfs2_releasepage(struct page *page, gfp_t wait) { if (!page_has_buffers(page)) @@ -658,363 +506,6 @@ static int ocfs2_releasepage(struct page *page, gfp_t wait) return try_to_free_buffers(page); } -static int ocfs2_is_overwrite(struct ocfs2_super *osb, - struct inode *inode, loff_t offset) -{ - int ret = 0; - u32 v_cpos = 0; - u32 p_cpos = 0; - unsigned int num_clusters = 0; - unsigned int ext_flags = 0; - - v_cpos = ocfs2_bytes_to_clusters(osb->sb, offset); - ret = ocfs2_get_clusters(inode, v_cpos, &p_cpos, - &num_clusters, &ext_flags); - if (ret < 0) { - mlog_errno(ret); - return ret; - } - - if (p_cpos && !(ext_flags & OCFS2_EXT_UNWRITTEN)) - return 1; - - return 0; -} - -static int ocfs2_direct_IO_zero_extend(struct ocfs2_super *osb, - struct inode *inode, loff_t offset, - u64 zero_len, int cluster_align) -{ - u32 p_cpos = 0; - u32 v_cpos = ocfs2_bytes_to_clusters(osb->sb, i_size_read(inode)); - unsigned int num_clusters = 0; - unsigned int ext_flags = 0; - int ret = 0; - - if (offset <= i_size_read(inode) || cluster_align) - return 0; - - ret = ocfs2_get_clusters(inode, v_cpos, &p_cpos, &num_clusters, - &ext_flags); - if (ret < 0) { - mlog_errno(ret); - return ret; - } - - if (p_cpos && !(ext_flags & OCFS2_EXT_UNWRITTEN)) { - u64 s = i_size_read(inode); - sector_t sector = ((u64)p_cpos << (osb->s_clustersize_bits - 9)) + - (do_div(s, osb->s_clustersize) >> 9); - - ret = blkdev_issue_zeroout(osb->sb->s_bdev, sector, - zero_len >> 9, GFP_NOFS, false); - if (ret < 0) - mlog_errno(ret); - } - - return ret; -} - -static int ocfs2_direct_IO_extend_no_holes(struct ocfs2_super *osb, - struct inode *inode, loff_t offset) -{ - u64 zero_start, zero_len, total_zero_len; - u32 p_cpos = 0, clusters_to_add; - u32 v_cpos = ocfs2_bytes_to_clusters(osb->sb, i_size_read(inode)); - unsigned int num_clusters = 0; - unsigned int ext_flags = 0; - u32 size_div, offset_div; - int ret = 0; - - { - u64 o = offset; - u64 s = i_size_read(inode); - - offset_div = do_div(o, osb->s_clustersize); - size_div = do_div(s, osb->s_clustersize); - } - - if (offset <= i_size_read(inode)) - return 0; - - clusters_to_add = ocfs2_bytes_to_clusters(inode->i_sb, offset) - - ocfs2_bytes_to_clusters(inode->i_sb, i_size_read(inode)); - total_zero_len = offset - i_size_read(inode); - if (clusters_to_add) - total_zero_len -= offset_div; - - /* Allocate clusters to fill out holes, and this is only needed - * when we add more than one clusters. Otherwise the cluster will - * be allocated during direct IO */ - if (clusters_to_add > 1) { - ret = ocfs2_extend_allocation(inode, - OCFS2_I(inode)->ip_clusters, - clusters_to_add - 1, 0); - if (ret) { - mlog_errno(ret); - goto out; - } - } - - while (total_zero_len) { - ret = ocfs2_get_clusters(inode, v_cpos, &p_cpos, &num_clusters, - &ext_flags); - if (ret < 0) { - mlog_errno(ret); - goto out; - } - - zero_start = ocfs2_clusters_to_bytes(osb->sb, p_cpos) + - size_div; - zero_len = ocfs2_clusters_to_bytes(osb->sb, num_clusters) - - size_div; - zero_len = min(total_zero_len, zero_len); - - if (p_cpos && !(ext_flags & OCFS2_EXT_UNWRITTEN)) { - ret = blkdev_issue_zeroout(osb->sb->s_bdev, - zero_start >> 9, zero_len >> 9, - GFP_NOFS, false); - if (ret < 0) { - mlog_errno(ret); - goto out; - } - } - - total_zero_len -= zero_len; - v_cpos += ocfs2_bytes_to_clusters(osb->sb, zero_len + size_div); - - /* Only at first iteration can be cluster not aligned. - * So set size_div to 0 for the rest */ - size_div = 0; - } - -out: - return ret; -} - -static ssize_t ocfs2_direct_IO_write(struct kiocb *iocb, - struct iov_iter *iter, - loff_t offset) -{ - ssize_t ret = 0; - ssize_t written = 0; - bool orphaned = false; - int is_overwrite = 0; - struct file *file = iocb->ki_filp; - struct inode *inode = file_inode(file)->i_mapping->host; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - struct buffer_head *di_bh = NULL; - size_t count = iter->count; - journal_t *journal = osb->journal->j_journal; - u64 zero_len_head, zero_len_tail; - int cluster_align_head, cluster_align_tail; - loff_t final_size = offset + count; - int append_write = offset >= i_size_read(inode) ? 1 : 0; - unsigned int num_clusters = 0; - unsigned int ext_flags = 0; - - { - u64 o = offset; - u64 s = i_size_read(inode); - - zero_len_head = do_div(o, 1 << osb->s_clustersize_bits); - cluster_align_head = !zero_len_head; - - zero_len_tail = osb->s_clustersize - - do_div(s, osb->s_clustersize); - if ((offset - i_size_read(inode)) < zero_len_tail) - zero_len_tail = offset - i_size_read(inode); - cluster_align_tail = !zero_len_tail; - } - - /* - * when final_size > inode->i_size, inode->i_size will be - * updated after direct write, so add the inode to orphan - * dir first. - */ - if (final_size > i_size_read(inode)) { - ret = ocfs2_add_inode_to_orphan(osb, inode); - if (ret < 0) { - mlog_errno(ret); - goto out; - } - orphaned = true; - } - - if (append_write) { - ret = ocfs2_inode_lock(inode, NULL, 1); - if (ret < 0) { - mlog_errno(ret); - goto clean_orphan; - } - - /* zeroing out the previously allocated cluster tail - * that but not zeroed */ - if (ocfs2_sparse_alloc(OCFS2_SB(inode->i_sb))) { - down_read(&OCFS2_I(inode)->ip_alloc_sem); - ret = ocfs2_direct_IO_zero_extend(osb, inode, offset, - zero_len_tail, cluster_align_tail); - up_read(&OCFS2_I(inode)->ip_alloc_sem); - } else { - down_write(&OCFS2_I(inode)->ip_alloc_sem); - ret = ocfs2_direct_IO_extend_no_holes(osb, inode, - offset); - up_write(&OCFS2_I(inode)->ip_alloc_sem); - } - if (ret < 0) { - mlog_errno(ret); - ocfs2_inode_unlock(inode, 1); - goto clean_orphan; - } - - is_overwrite = ocfs2_is_overwrite(osb, inode, offset); - if (is_overwrite < 0) { - mlog_errno(is_overwrite); - ret = is_overwrite; - ocfs2_inode_unlock(inode, 1); - goto clean_orphan; - } - - ocfs2_inode_unlock(inode, 1); - } - - written = __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev, iter, - offset, ocfs2_direct_IO_get_blocks, - ocfs2_dio_end_io, NULL, 0); - /* overwrite aio may return -EIOCBQUEUED, and it is not an error */ - if ((written < 0) && (written != -EIOCBQUEUED)) { - loff_t i_size = i_size_read(inode); - - if (offset + count > i_size) { - ret = ocfs2_inode_lock(inode, &di_bh, 1); - if (ret < 0) { - mlog_errno(ret); - goto clean_orphan; - } - - if (i_size == i_size_read(inode)) { - ret = ocfs2_truncate_file(inode, di_bh, - i_size); - if (ret < 0) { - if (ret != -ENOSPC) - mlog_errno(ret); - - ocfs2_inode_unlock(inode, 1); - brelse(di_bh); - di_bh = NULL; - goto clean_orphan; - } - } - - ocfs2_inode_unlock(inode, 1); - brelse(di_bh); - di_bh = NULL; - - ret = jbd2_journal_force_commit(journal); - if (ret < 0) - mlog_errno(ret); - } - } else if (written > 0 && append_write && !is_overwrite && - !cluster_align_head) { - /* zeroing out the allocated cluster head */ - u32 p_cpos = 0; - u32 v_cpos = ocfs2_bytes_to_clusters(osb->sb, offset); - - ret = ocfs2_inode_lock(inode, NULL, 0); - if (ret < 0) { - mlog_errno(ret); - goto clean_orphan; - } - - ret = ocfs2_get_clusters(inode, v_cpos, &p_cpos, - &num_clusters, &ext_flags); - if (ret < 0) { - mlog_errno(ret); - ocfs2_inode_unlock(inode, 0); - goto clean_orphan; - } - - BUG_ON(!p_cpos || (ext_flags & OCFS2_EXT_UNWRITTEN)); - - ret = blkdev_issue_zeroout(osb->sb->s_bdev, - (u64)p_cpos << (osb->s_clustersize_bits - 9), - zero_len_head >> 9, GFP_NOFS, false); - if (ret < 0) - mlog_errno(ret); - - ocfs2_inode_unlock(inode, 0); - } - -clean_orphan: - if (orphaned) { - int tmp_ret; - int update_isize = written > 0 ? 1 : 0; - loff_t end = update_isize ? offset + written : 0; - - tmp_ret = ocfs2_inode_lock(inode, &di_bh, 1); - if (tmp_ret < 0) { - ret = tmp_ret; - mlog_errno(ret); - goto out; - } - - tmp_ret = ocfs2_del_inode_from_orphan(osb, inode, di_bh, - update_isize, end); - if (tmp_ret < 0) { - ocfs2_inode_unlock(inode, 1); - ret = tmp_ret; - mlog_errno(ret); - brelse(di_bh); - goto out; - } - - ocfs2_inode_unlock(inode, 1); - brelse(di_bh); - - tmp_ret = jbd2_journal_force_commit(journal); - if (tmp_ret < 0) { - ret = tmp_ret; - mlog_errno(tmp_ret); - } - } - -out: - if (ret >= 0) - ret = written; - return ret; -} - -static ssize_t ocfs2_direct_IO(struct kiocb *iocb, struct iov_iter *iter, - loff_t offset) -{ - struct file *file = iocb->ki_filp; - struct inode *inode = file_inode(file)->i_mapping->host; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - int full_coherency = !(osb->s_mount_opt & - OCFS2_MOUNT_COHERENCY_BUFFERED); - - /* - * Fallback to buffered I/O if we see an inode without - * extents. - */ - if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) - return 0; - - /* Fallback to buffered I/O if we are appending and - * concurrent O_DIRECT writes are allowed. - */ - if (i_size_read(inode) <= offset && !full_coherency) - return 0; - - if (iov_iter_rw(iter) == READ) - return __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev, - iter, offset, - ocfs2_direct_IO_get_blocks, - ocfs2_dio_end_io, NULL, 0); - else - return ocfs2_direct_IO_write(iocb, iter, offset); -} - static void ocfs2_figure_cluster_boundaries(struct ocfs2_super *osb, u32 cpos, unsigned int *start, @@ -1201,6 +692,13 @@ next_bh: #define OCFS2_MAX_CLUSTERS_PER_PAGE (PAGE_CACHE_SIZE / OCFS2_MIN_CLUSTERSIZE) +struct ocfs2_unwritten_extent { + struct list_head ue_node; + struct list_head ue_ip_node; + u32 ue_cpos; + u32 ue_phys; +}; + /* * Describe the state of a single cluster to be written to. */ @@ -1212,7 +710,7 @@ struct ocfs2_write_cluster_desc { * filled. */ unsigned c_new; - unsigned c_unwritten; + unsigned c_clear_unwritten; unsigned c_needs_zero; }; @@ -1224,6 +722,9 @@ struct ocfs2_write_ctxt { /* First cluster allocated in a nonsparse extend */ u32 w_first_new_cpos; + /* Type of caller. Must be one of buffer, mmap, direct. */ + ocfs2_write_type_t w_type; + struct ocfs2_write_cluster_desc w_desc[OCFS2_MAX_CLUSTERS_PER_PAGE]; /* @@ -1272,6 +773,8 @@ struct ocfs2_write_ctxt { struct buffer_head *w_di_bh; struct ocfs2_cached_dealloc_ctxt w_dealloc; + + struct list_head w_unwritten_list; }; void ocfs2_unlock_and_free_pages(struct page **pages, int num_pages) @@ -1310,8 +813,25 @@ static void ocfs2_unlock_pages(struct ocfs2_write_ctxt *wc) ocfs2_unlock_and_free_pages(wc->w_pages, wc->w_num_pages); } -static void ocfs2_free_write_ctxt(struct ocfs2_write_ctxt *wc) +static void ocfs2_free_unwritten_list(struct inode *inode, + struct list_head *head) { + struct ocfs2_inode_info *oi = OCFS2_I(inode); + struct ocfs2_unwritten_extent *ue = NULL, *tmp = NULL; + + list_for_each_entry_safe(ue, tmp, head, ue_node) { + list_del(&ue->ue_node); + spin_lock(&oi->ip_lock); + list_del(&ue->ue_ip_node); + spin_unlock(&oi->ip_lock); + kfree(ue); + } +} + +static void ocfs2_free_write_ctxt(struct inode *inode, + struct ocfs2_write_ctxt *wc) +{ + ocfs2_free_unwritten_list(inode, &wc->w_unwritten_list); ocfs2_unlock_pages(wc); brelse(wc->w_di_bh); kfree(wc); @@ -1319,7 +839,8 @@ static void ocfs2_free_write_ctxt(struct ocfs2_write_ctxt *wc) static int ocfs2_alloc_write_ctxt(struct ocfs2_write_ctxt **wcp, struct ocfs2_super *osb, loff_t pos, - unsigned len, struct buffer_head *di_bh) + unsigned len, ocfs2_write_type_t type, + struct buffer_head *di_bh) { u32 cend; struct ocfs2_write_ctxt *wc; @@ -1334,6 +855,7 @@ static int ocfs2_alloc_write_ctxt(struct ocfs2_write_ctxt **wcp, wc->w_clen = cend - wc->w_cpos + 1; get_bh(di_bh); wc->w_di_bh = di_bh; + wc->w_type = type; if (unlikely(PAGE_CACHE_SHIFT > osb->s_clustersize_bits)) wc->w_large_pages = 1; @@ -1341,6 +863,7 @@ static int ocfs2_alloc_write_ctxt(struct ocfs2_write_ctxt **wcp, wc->w_large_pages = 0; ocfs2_init_dealloc_ctxt(&wc->w_dealloc); + INIT_LIST_HEAD(&wc->w_unwritten_list); *wcp = wc; @@ -1401,12 +924,13 @@ static void ocfs2_write_failure(struct inode *inode, to = user_pos + user_len; struct page *tmppage; - ocfs2_zero_new_buffers(wc->w_target_page, from, to); + if (wc->w_target_page) + ocfs2_zero_new_buffers(wc->w_target_page, from, to); for(i = 0; i < wc->w_num_pages; i++) { tmppage = wc->w_pages[i]; - if (page_has_buffers(tmppage)) { + if (tmppage && page_has_buffers(tmppage)) { if (ocfs2_should_order_data(inode)) ocfs2_jbd2_file_inode(wc->w_handle, inode); @@ -1536,11 +1060,13 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping, wc->w_num_pages = 1; start = target_index; } + end_index = (user_pos + user_len - 1) >> PAGE_CACHE_SHIFT; for(i = 0; i < wc->w_num_pages; i++) { index = start + i; - if (index == target_index && mmap_page) { + if (index >= target_index && index <= end_index && + wc->w_type == OCFS2_WRITE_MMAP) { /* * ocfs2_pagemkwrite() is a little different * and wants us to directly use the page @@ -1559,6 +1085,11 @@ static int ocfs2_grab_pages_for_write(struct address_space *mapping, page_cache_get(mmap_page); wc->w_pages[i] = mmap_page; wc->w_target_locked = true; + } else if (index >= target_index && index <= end_index && + wc->w_type == OCFS2_WRITE_DIRECT) { + /* Direct write has no mapping page. */ + wc->w_pages[i] = NULL; + continue; } else { wc->w_pages[i] = find_or_create_page(mapping, index, GFP_NOFS); @@ -1583,19 +1114,20 @@ out: * Prepare a single cluster for write one cluster into the file. */ static int ocfs2_write_cluster(struct address_space *mapping, - u32 phys, unsigned int unwritten, + u32 *phys, unsigned int new, + unsigned int clear_unwritten, unsigned int should_zero, struct ocfs2_alloc_context *data_ac, struct ocfs2_alloc_context *meta_ac, struct ocfs2_write_ctxt *wc, u32 cpos, loff_t user_pos, unsigned user_len) { - int ret, i, new; - u64 v_blkno, p_blkno; + int ret, i; + u64 p_blkno; struct inode *inode = mapping->host; struct ocfs2_extent_tree et; + int bpc = ocfs2_clusters_to_blocks(inode->i_sb, 1); - new = phys == 0 ? 1 : 0; if (new) { u32 tmp_pos; @@ -1605,9 +1137,9 @@ static int ocfs2_write_cluster(struct address_space *mapping, */ tmp_pos = cpos; ret = ocfs2_add_inode_data(OCFS2_SB(inode->i_sb), inode, - &tmp_pos, 1, 0, wc->w_di_bh, - wc->w_handle, data_ac, - meta_ac, NULL); + &tmp_pos, 1, !clear_unwritten, + wc->w_di_bh, wc->w_handle, + data_ac, meta_ac, NULL); /* * This shouldn't happen because we must have already * calculated the correct meta data allocation required. The @@ -1624,11 +1156,11 @@ static int ocfs2_write_cluster(struct address_space *mapping, mlog_errno(ret); goto out; } - } else if (unwritten) { + } else if (clear_unwritten) { ocfs2_init_dinode_extent_tree(&et, INODE_CACHE(inode), wc->w_di_bh); ret = ocfs2_mark_extent_written(inode, &et, - wc->w_handle, cpos, 1, phys, + wc->w_handle, cpos, 1, *phys, meta_ac, &wc->w_dealloc); if (ret < 0) { mlog_errno(ret); @@ -1636,30 +1168,33 @@ static int ocfs2_write_cluster(struct address_space *mapping, } } - if (should_zero) - v_blkno = ocfs2_clusters_to_blocks(inode->i_sb, cpos); - else - v_blkno = user_pos >> inode->i_sb->s_blocksize_bits; - /* * The only reason this should fail is due to an inability to * find the extent added. */ - ret = ocfs2_extent_map_get_blocks(inode, v_blkno, &p_blkno, NULL, - NULL); + ret = ocfs2_get_clusters(inode, cpos, phys, NULL, NULL); if (ret < 0) { mlog(ML_ERROR, "Get physical blkno failed for inode %llu, " - "at logical block %llu", - (unsigned long long)OCFS2_I(inode)->ip_blkno, - (unsigned long long)v_blkno); + "at logical cluster %u", + (unsigned long long)OCFS2_I(inode)->ip_blkno, cpos); goto out; } - BUG_ON(p_blkno == 0); + BUG_ON(*phys == 0); + + p_blkno = ocfs2_clusters_to_blocks(inode->i_sb, *phys); + if (!should_zero) + p_blkno += (user_pos >> inode->i_sb->s_blocksize_bits) & (u64)(bpc - 1); for(i = 0; i < wc->w_num_pages; i++) { int tmpret; + /* This is the direct io target page. */ + if (wc->w_pages[i] == NULL) { + p_blkno++; + continue; + } + tmpret = ocfs2_prepare_page_for_write(inode, &p_blkno, wc, wc->w_pages[i], cpos, user_pos, user_len, @@ -1706,8 +1241,9 @@ static int ocfs2_write_cluster_by_desc(struct address_space *mapping, if ((cluster_off + local_len) > osb->s_clustersize) local_len = osb->s_clustersize - cluster_off; - ret = ocfs2_write_cluster(mapping, desc->c_phys, - desc->c_unwritten, + ret = ocfs2_write_cluster(mapping, &desc->c_phys, + desc->c_new, + desc->c_clear_unwritten, desc->c_needs_zero, data_ac, meta_ac, wc, desc->c_cpos, pos, local_len); @@ -1778,6 +1314,66 @@ static void ocfs2_set_target_boundaries(struct ocfs2_super *osb, } /* + * Check if this extent is marked UNWRITTEN by direct io. If so, we need not to + * do the zero work. And should not to clear UNWRITTEN since it will be cleared + * by the direct io procedure. + * If this is a new extent that allocated by direct io, we should mark it in + * the ip_unwritten_list. + */ +static int ocfs2_unwritten_check(struct inode *inode, + struct ocfs2_write_ctxt *wc, + struct ocfs2_write_cluster_desc *desc) +{ + struct ocfs2_inode_info *oi = OCFS2_I(inode); + struct ocfs2_unwritten_extent *ue = NULL, *new = NULL; + int ret = 0; + + if (!desc->c_needs_zero) + return 0; + +retry: + spin_lock(&oi->ip_lock); + /* Needs not to zero no metter buffer or direct. The one who is zero + * the cluster is doing zero. And he will clear unwritten after all + * cluster io finished. */ + list_for_each_entry(ue, &oi->ip_unwritten_list, ue_ip_node) { + if (desc->c_cpos == ue->ue_cpos) { + BUG_ON(desc->c_new); + desc->c_needs_zero = 0; + desc->c_clear_unwritten = 0; + goto unlock; + } + } + + if (wc->w_type != OCFS2_WRITE_DIRECT) + goto unlock; + + if (new == NULL) { + spin_unlock(&oi->ip_lock); + new = kmalloc(sizeof(struct ocfs2_unwritten_extent), + GFP_NOFS); + if (new == NULL) { + ret = -ENOMEM; + goto out; + } + goto retry; + } + /* This direct write will doing zero. */ + new->ue_cpos = desc->c_cpos; + new->ue_phys = desc->c_phys; + desc->c_clear_unwritten = 0; + list_add_tail(&new->ue_ip_node, &oi->ip_unwritten_list); + list_add_tail(&new->ue_node, &wc->w_unwritten_list); + new = NULL; +unlock: + spin_unlock(&oi->ip_lock); +out: + if (new) + kfree(new); + return ret; +} + +/* * Populate each single-cluster write descriptor in the write context * with information about the i/o to be done. * @@ -1852,14 +1448,21 @@ static int ocfs2_populate_write_desc(struct inode *inode, if (phys == 0) { desc->c_new = 1; desc->c_needs_zero = 1; + desc->c_clear_unwritten = 1; *clusters_to_alloc = *clusters_to_alloc + 1; } if (ext_flags & OCFS2_EXT_UNWRITTEN) { - desc->c_unwritten = 1; + desc->c_clear_unwritten = 1; desc->c_needs_zero = 1; } + ret = ocfs2_unwritten_check(inode, wc, desc); + if (ret) { + mlog_errno(ret); + goto out; + } + num_clusters--; } @@ -2022,8 +1625,10 @@ static int ocfs2_expand_nonsparse_inode(struct inode *inode, if (ret) mlog_errno(ret); - wc->w_first_new_cpos = - ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode)); + /* There is no wc if this is call from direct. */ + if (wc) + wc->w_first_new_cpos = + ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode)); return ret; } @@ -2077,9 +1682,8 @@ out: return ret; } -int ocfs2_write_begin_nolock(struct file *filp, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned flags, +int ocfs2_write_begin_nolock(struct address_space *mapping, + loff_t pos, unsigned len, ocfs2_write_type_t type, struct page **pagep, void **fsdata, struct buffer_head *di_bh, struct page *mmap_page) { @@ -2096,7 +1700,7 @@ int ocfs2_write_begin_nolock(struct file *filp, int try_free = 1, ret1; try_again: - ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, di_bh); + ret = ocfs2_alloc_write_ctxt(&wc, osb, pos, len, type, di_bh); if (ret) { mlog_errno(ret); return ret; @@ -2115,14 +1719,17 @@ try_again: } } - if (ocfs2_sparse_alloc(osb)) - ret = ocfs2_zero_tail(inode, di_bh, pos); - else - ret = ocfs2_expand_nonsparse_inode(inode, di_bh, pos, len, - wc); - if (ret) { - mlog_errno(ret); - goto out; + /* Direct io change i_size late, should not zero tail here. */ + if (type != OCFS2_WRITE_DIRECT) { + if (ocfs2_sparse_alloc(osb)) + ret = ocfs2_zero_tail(inode, di_bh, pos); + else + ret = ocfs2_expand_nonsparse_inode(inode, di_bh, pos, + len, wc); + if (ret) { + mlog_errno(ret); + goto out; + } } ret = ocfs2_check_range_for_refcount(inode, pos, len); @@ -2153,7 +1760,7 @@ try_again: (unsigned long long)OCFS2_I(inode)->ip_blkno, (long long)i_size_read(inode), le32_to_cpu(di->i_clusters), - pos, len, flags, mmap_page, + pos, len, type, mmap_page, clusters_to_alloc, extents_to_split); /* @@ -2183,17 +1790,17 @@ try_again: credits = ocfs2_calc_extend_credits(inode->i_sb, &di->id2.i_list); - - } + } else if (type == OCFS2_WRITE_DIRECT) + /* direct write needs not to start trans if no extents alloc. */ + goto success; /* * We have to zero sparse allocated clusters, unwritten extent clusters, * and non-sparse clusters we just extended. For non-sparse writes, * we know zeros will only be needed in the first and/or last cluster. */ - if (clusters_to_alloc || extents_to_split || - (wc->w_clen && (wc->w_desc[0].c_needs_zero || - wc->w_desc[wc->w_clen - 1].c_needs_zero))) + if (wc->w_clen && (wc->w_desc[0].c_needs_zero || + wc->w_desc[wc->w_clen - 1].c_needs_zero)) cluster_of_pages = 1; else cluster_of_pages = 0; @@ -2260,7 +1867,8 @@ try_again: ocfs2_free_alloc_context(meta_ac); success: - *pagep = wc->w_target_page; + if (pagep) + *pagep = wc->w_target_page; *fsdata = wc; return 0; out_quota: @@ -2271,7 +1879,7 @@ out_commit: ocfs2_commit_trans(osb, handle); out: - ocfs2_free_write_ctxt(wc); + ocfs2_free_write_ctxt(inode, wc); if (data_ac) { ocfs2_free_alloc_context(data_ac); @@ -2323,8 +1931,8 @@ static int ocfs2_write_begin(struct file *file, struct address_space *mapping, */ down_write(&OCFS2_I(inode)->ip_alloc_sem); - ret = ocfs2_write_begin_nolock(file, mapping, pos, len, flags, pagep, - fsdata, di_bh, NULL); + ret = ocfs2_write_begin_nolock(mapping, pos, len, OCFS2_WRITE_BUFFER, + pagep, fsdata, di_bh, NULL); if (ret) { mlog_errno(ret); goto out_fail; @@ -2381,12 +1989,16 @@ int ocfs2_write_end_nolock(struct address_space *mapping, handle_t *handle = wc->w_handle; struct page *tmppage; - ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), wc->w_di_bh, - OCFS2_JOURNAL_ACCESS_WRITE); - if (ret) { - copied = ret; - mlog_errno(ret); - goto out; + BUG_ON(!list_empty(&wc->w_unwritten_list)); + + if (handle) { + ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), + wc->w_di_bh, OCFS2_JOURNAL_ACCESS_WRITE); + if (ret) { + copied = ret; + mlog_errno(ret); + goto out; + } } if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) { @@ -2394,18 +2006,23 @@ int ocfs2_write_end_nolock(struct address_space *mapping, goto out_write_size; } - if (unlikely(copied < len)) { + if (unlikely(copied < len) && wc->w_target_page) { if (!PageUptodate(wc->w_target_page)) copied = 0; ocfs2_zero_new_buffers(wc->w_target_page, start+copied, start+len); } - flush_dcache_page(wc->w_target_page); + if (wc->w_target_page) + flush_dcache_page(wc->w_target_page); for(i = 0; i < wc->w_num_pages; i++) { tmppage = wc->w_pages[i]; + /* This is the direct io target page. */ + if (tmppage == NULL) + continue; + if (tmppage == wc->w_target_page) { from = wc->w_target_from; to = wc->w_target_to; @@ -2424,25 +2041,29 @@ int ocfs2_write_end_nolock(struct address_space *mapping, } if (page_has_buffers(tmppage)) { - if (ocfs2_should_order_data(inode)) - ocfs2_jbd2_file_inode(wc->w_handle, inode); + if (handle && ocfs2_should_order_data(inode)) + ocfs2_jbd2_file_inode(handle, inode); block_commit_write(tmppage, from, to); } } out_write_size: - pos += copied; - if (pos > i_size_read(inode)) { - i_size_write(inode, pos); - mark_inode_dirty(inode); - } - inode->i_blocks = ocfs2_inode_sector_count(inode); - di->i_size = cpu_to_le64((u64)i_size_read(inode)); - inode->i_mtime = inode->i_ctime = CURRENT_TIME; - di->i_mtime = di->i_ctime = cpu_to_le64(inode->i_mtime.tv_sec); - di->i_mtime_nsec = di->i_ctime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec); - ocfs2_update_inode_fsync_trans(handle, inode, 1); - ocfs2_journal_dirty(handle, wc->w_di_bh); + /* Direct io do not update i_size here. */ + if (wc->w_type != OCFS2_WRITE_DIRECT) { + pos += copied; + if (pos > i_size_read(inode)) { + i_size_write(inode, pos); + mark_inode_dirty(inode); + } + inode->i_blocks = ocfs2_inode_sector_count(inode); + di->i_size = cpu_to_le64((u64)i_size_read(inode)); + inode->i_mtime = inode->i_ctime = CURRENT_TIME; + di->i_mtime = di->i_ctime = cpu_to_le64(inode->i_mtime.tv_sec); + di->i_mtime_nsec = di->i_ctime_nsec = cpu_to_le32(inode->i_mtime.tv_nsec); + ocfs2_update_inode_fsync_trans(handle, inode, 1); + } + if (handle) + ocfs2_journal_dirty(handle, wc->w_di_bh); out: /* unlock pages before dealloc since it needs acquiring j_trans_barrier @@ -2452,7 +2073,8 @@ out: */ ocfs2_unlock_pages(wc); - ocfs2_commit_trans(osb, handle); + if (handle) + ocfs2_commit_trans(osb, handle); ocfs2_run_deallocs(osb, &wc->w_dealloc); @@ -2477,6 +2099,360 @@ static int ocfs2_write_end(struct file *file, struct address_space *mapping, return ret; } +struct ocfs2_dio_write_ctxt { + struct list_head dw_zero_list; + unsigned dw_zero_count; + int dw_orphaned; + pid_t dw_writer_pid; +}; + +static struct ocfs2_dio_write_ctxt * +ocfs2_dio_alloc_write_ctx(struct buffer_head *bh, int *alloc) +{ + struct ocfs2_dio_write_ctxt *dwc = NULL; + + if (bh->b_private) + return bh->b_private; + + dwc = kmalloc(sizeof(struct ocfs2_dio_write_ctxt), GFP_NOFS); + if (dwc == NULL) + return NULL; + INIT_LIST_HEAD(&dwc->dw_zero_list); + dwc->dw_zero_count = 0; + dwc->dw_orphaned = 0; + dwc->dw_writer_pid = task_pid_nr(current); + bh->b_private = dwc; + *alloc = 1; + + return dwc; +} + +static void ocfs2_dio_free_write_ctx(struct inode *inode, + struct ocfs2_dio_write_ctxt *dwc) +{ + ocfs2_free_unwritten_list(inode, &dwc->dw_zero_list); + kfree(dwc); +} + +/* + * TODO: Make this into a generic get_blocks function. + * + * From do_direct_io in direct-io.c: + * "So what we do is to permit the ->get_blocks function to populate + * bh.b_size with the size of IO which is permitted at this offset and + * this i_blkbits." + * + * This function is called directly from get_more_blocks in direct-io.c. + * + * called like this: dio->get_blocks(dio->inode, fs_startblk, + * fs_count, map_bh, dio->rw == WRITE); + */ +static int ocfs2_dio_get_block(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create) +{ + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct ocfs2_inode_info *oi = OCFS2_I(inode); + struct ocfs2_write_ctxt *wc; + struct ocfs2_write_cluster_desc *desc = NULL; + struct ocfs2_dio_write_ctxt *dwc = NULL; + struct buffer_head *di_bh = NULL; + u64 p_blkno; + loff_t pos = iblock << inode->i_sb->s_blocksize_bits; + unsigned len, total_len = bh_result->b_size; + int ret = 0, first_get_block = 0; + + len = osb->s_clustersize - (pos & (osb->s_clustersize - 1)); + len = min(total_len, len); + + mlog(0, "get block of %lu at %llu:%u req %u\n", + inode->i_ino, pos, len, total_len); + + /* + * Because we need to change file size in ocfs2_dio_end_io_write(), or + * we may need to add it to orphan dir. So can not fall to fast path + * while file size will be changed. + */ + if (pos + total_len <= i_size_read(inode)) { + down_read(&oi->ip_alloc_sem); + /* This is the fast path for re-write. */ + ret = ocfs2_get_block(inode, iblock, bh_result, create); + + up_read(&oi->ip_alloc_sem); + + if (buffer_mapped(bh_result) && + !buffer_new(bh_result) && + ret == 0) + goto out; + + /* Clear state set by ocfs2_get_block. */ + bh_result->b_state = 0; + } + + dwc = ocfs2_dio_alloc_write_ctx(bh_result, &first_get_block); + if (unlikely(dwc == NULL)) { + ret = -ENOMEM; + mlog_errno(ret); + goto out; + } + + if (ocfs2_clusters_for_bytes(inode->i_sb, pos + total_len) > + ocfs2_clusters_for_bytes(inode->i_sb, i_size_read(inode)) && + !dwc->dw_orphaned) { + /* + * when we are going to alloc extents beyond file size, add the + * inode to orphan dir, so we can recall those spaces when + * system crashed during write. + */ + ret = ocfs2_add_inode_to_orphan(osb, inode); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + dwc->dw_orphaned = 1; + } + + ret = ocfs2_inode_lock(inode, &di_bh, 1); + if (ret) { + mlog_errno(ret); + goto out; + } + + down_write(&oi->ip_alloc_sem); + + if (first_get_block) { + if (ocfs2_sparse_alloc(OCFS2_SB(inode->i_sb))) + ret = ocfs2_zero_tail(inode, di_bh, pos); + else + ret = ocfs2_expand_nonsparse_inode(inode, di_bh, pos, + total_len, NULL); + if (ret < 0) { + mlog_errno(ret); + goto unlock; + } + } + + ret = ocfs2_write_begin_nolock(inode->i_mapping, pos, len, + OCFS2_WRITE_DIRECT, NULL, + (void **)&wc, di_bh, NULL); + if (ret) { + mlog_errno(ret); + goto unlock; + } + + desc = &wc->w_desc[0]; + + p_blkno = ocfs2_clusters_to_blocks(inode->i_sb, desc->c_phys); + BUG_ON(p_blkno == 0); + p_blkno += iblock & (u64)(ocfs2_clusters_to_blocks(inode->i_sb, 1) - 1); + + map_bh(bh_result, inode->i_sb, p_blkno); + bh_result->b_size = len; + if (desc->c_needs_zero) + set_buffer_new(bh_result); + + /* May sleep in end_io. It should not happen in a irq context. So defer + * it to dio work queue. */ + set_buffer_defer_completion(bh_result); + + if (!list_empty(&wc->w_unwritten_list)) { + struct ocfs2_unwritten_extent *ue = NULL; + + ue = list_first_entry(&wc->w_unwritten_list, + struct ocfs2_unwritten_extent, + ue_node); + BUG_ON(ue->ue_cpos != desc->c_cpos); + /* The physical address may be 0, fill it. */ + ue->ue_phys = desc->c_phys; + + list_splice_tail_init(&wc->w_unwritten_list, &dwc->dw_zero_list); + dwc->dw_zero_count++; + } + + ret = ocfs2_write_end_nolock(inode->i_mapping, pos, len, len, NULL, wc); + BUG_ON(ret != len); + ret = 0; +unlock: + up_write(&oi->ip_alloc_sem); + ocfs2_inode_unlock(inode, 1); + brelse(di_bh); +out: + if (ret < 0) + ret = -EIO; + return ret; +} + +static void ocfs2_dio_end_io_write(struct inode *inode, + struct ocfs2_dio_write_ctxt *dwc, + loff_t offset, + ssize_t bytes) +{ + struct ocfs2_cached_dealloc_ctxt dealloc; + struct ocfs2_extent_tree et; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct ocfs2_inode_info *oi = OCFS2_I(inode); + struct ocfs2_unwritten_extent *ue = NULL; + struct buffer_head *di_bh = NULL; + struct ocfs2_dinode *di; + struct ocfs2_alloc_context *data_ac = NULL; + struct ocfs2_alloc_context *meta_ac = NULL; + handle_t *handle = NULL; + loff_t end = offset + bytes; + int ret = 0, credits = 0, locked = 0; + + ocfs2_init_dealloc_ctxt(&dealloc); + + /* We do clear unwritten, delete orphan, change i_size here. If neither + * of these happen, we can skip all this. */ + if (list_empty(&dwc->dw_zero_list) && + end <= i_size_read(inode) && + !dwc->dw_orphaned) + goto out; + + /* ocfs2_file_write_iter will get i_mutex, so we need not lock if we + * are in that context. */ + if (dwc->dw_writer_pid != task_pid_nr(current)) { + mutex_lock(&inode->i_mutex); + locked = 1; + } + + ret = ocfs2_inode_lock(inode, &di_bh, 1); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + down_write(&oi->ip_alloc_sem); + + /* Delete orphan before acquire i_mutex. */ + if (dwc->dw_orphaned) { + BUG_ON(dwc->dw_writer_pid != task_pid_nr(current)); + + end = end > i_size_read(inode) ? end : 0; + + ret = ocfs2_del_inode_from_orphan(osb, inode, di_bh, + !!end, end); + if (ret < 0) + mlog_errno(ret); + } + + di = (struct ocfs2_dinode *)di_bh; + + ocfs2_init_dinode_extent_tree(&et, INODE_CACHE(inode), di_bh); + + ret = ocfs2_lock_allocators(inode, &et, 0, dwc->dw_zero_count*2, + &data_ac, &meta_ac); + if (ret) { + mlog_errno(ret); + goto unlock; + } + + credits = ocfs2_calc_extend_credits(inode->i_sb, &di->id2.i_list); + + handle = ocfs2_start_trans(osb, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + mlog_errno(ret); + goto unlock; + } + ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh, + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret) { + mlog_errno(ret); + goto commit; + } + + list_for_each_entry(ue, &dwc->dw_zero_list, ue_node) { + ret = ocfs2_mark_extent_written(inode, &et, handle, + ue->ue_cpos, 1, + ue->ue_phys, + meta_ac, &dealloc); + if (ret < 0) { + mlog_errno(ret); + break; + } + } + + if (end > i_size_read(inode)) { + ret = ocfs2_set_inode_size(handle, inode, di_bh, end); + if (ret < 0) + mlog_errno(ret); + } +commit: + ocfs2_commit_trans(osb, handle); +unlock: + up_write(&oi->ip_alloc_sem); + ocfs2_inode_unlock(inode, 1); + brelse(di_bh); +out: + if (data_ac) + ocfs2_free_alloc_context(data_ac); + if (meta_ac) + ocfs2_free_alloc_context(meta_ac); + ocfs2_run_deallocs(osb, &dealloc); + if (locked) + mutex_unlock(&inode->i_mutex); + ocfs2_dio_free_write_ctx(inode, dwc); +} + +/* + * ocfs2_dio_end_io is called by the dio core when a dio is finished. We're + * particularly interested in the aio/dio case. We use the rw_lock DLM lock + * to protect io on one node from truncation on another. + */ +static int ocfs2_dio_end_io(struct kiocb *iocb, + loff_t offset, + ssize_t bytes, + void *private) +{ + struct inode *inode = file_inode(iocb->ki_filp); + int level; + + if (bytes <= 0) + return 0; + + /* this io's submitter should not have unlocked this before we could */ + BUG_ON(!ocfs2_iocb_is_rw_locked(iocb)); + + if (private) + ocfs2_dio_end_io_write(inode, private, offset, bytes); + + ocfs2_iocb_clear_rw_locked(iocb); + + level = ocfs2_iocb_rw_locked_level(iocb); + ocfs2_rw_unlock(inode, level); + return 0; +} + +static ssize_t ocfs2_direct_IO(struct kiocb *iocb, struct iov_iter *iter, + loff_t offset) +{ + struct file *file = iocb->ki_filp; + struct inode *inode = file_inode(file)->i_mapping->host; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + loff_t end = offset + iter->count; + get_block_t *get_block; + + /* + * Fallback to buffered I/O if we see an inode without + * extents. + */ + if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) + return 0; + + /* Fallback to buffered I/O if we do not support append dio. */ + if (end > i_size_read(inode) && !ocfs2_supports_append_dio(osb)) + return 0; + + if (iov_iter_rw(iter) == READ) + get_block = ocfs2_get_block; + else + get_block = ocfs2_dio_get_block; + + return __blockdev_direct_IO(iocb, inode, inode->i_sb->s_bdev, + iter, offset, get_block, + ocfs2_dio_end_io, NULL, 0); +} + const struct address_space_operations ocfs2_aops = { .readpage = ocfs2_readpage, .readpages = ocfs2_readpages, diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h index 24e496d6bdcd..b1c9f28a57b1 100644 --- a/fs/ocfs2/aops.h +++ b/fs/ocfs2/aops.h @@ -47,9 +47,14 @@ int ocfs2_write_end_nolock(struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata); -int ocfs2_write_begin_nolock(struct file *filp, - struct address_space *mapping, - loff_t pos, unsigned len, unsigned flags, +typedef enum { + OCFS2_WRITE_BUFFER = 0, + OCFS2_WRITE_DIRECT, + OCFS2_WRITE_MMAP, +} ocfs2_write_type_t; + +int ocfs2_write_begin_nolock(struct address_space *mapping, + loff_t pos, unsigned len, ocfs2_write_type_t type, struct page **pagep, void **fsdata, struct buffer_head *di_bh, struct page *mmap_page); @@ -79,7 +84,6 @@ static inline void ocfs2_iocb_set_rw_locked(struct kiocb *iocb, int level) enum ocfs2_iocb_lock_bits { OCFS2_IOCB_RW_LOCK = 0, OCFS2_IOCB_RW_LOCK_LEVEL, - OCFS2_IOCB_UNALIGNED_IO, OCFS2_IOCB_NUM_LOCKS }; @@ -88,11 +92,4 @@ enum ocfs2_iocb_lock_bits { #define ocfs2_iocb_rw_locked_level(iocb) \ test_bit(OCFS2_IOCB_RW_LOCK_LEVEL, (unsigned long *)&iocb->private) -#define ocfs2_iocb_set_unaligned_aio(iocb) \ - set_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private) -#define ocfs2_iocb_clear_unaligned_aio(iocb) \ - clear_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private) -#define ocfs2_iocb_is_unaligned_aio(iocb) \ - test_bit(OCFS2_IOCB_UNALIGNED_IO, (unsigned long *)&iocb->private) - #endif /* OCFS2_FILE_H */ diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index ef6a2ec494de..bd15929b5f92 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -1444,8 +1444,8 @@ static void o2hb_region_release(struct config_item *item) debugfs_remove(reg->hr_debug_dir); kfree(reg->hr_db_livenodes); kfree(reg->hr_db_regnum); - kfree(reg->hr_debug_elapsed_time); - kfree(reg->hr_debug_pinned); + kfree(reg->hr_db_elapsed_time); + kfree(reg->hr_db_pinned); spin_lock(&o2hb_live_lock); list_del(®->hr_all_item); diff --git a/fs/ocfs2/dlm/dlmconvert.c b/fs/ocfs2/dlm/dlmconvert.c index e36d63ff1783..cdeafb4e7ed6 100644 --- a/fs/ocfs2/dlm/dlmconvert.c +++ b/fs/ocfs2/dlm/dlmconvert.c @@ -212,6 +212,12 @@ grant: if (lock->lksb->flags & DLM_LKSB_PUT_LVB) memcpy(res->lvb, lock->lksb->lvb, DLM_LVB_LEN); + /* + * Move the lock to the tail because it may be the only lock which has + * an invalid lvb. + */ + list_move_tail(&lock->list, &res->granted); + status = DLM_NORMAL; *call_ast = 1; goto unlock_exit; @@ -262,6 +268,7 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, struct dlm_lock *lock, int flags, int type) { enum dlm_status status; + u8 old_owner = res->owner; mlog(0, "type=%d, convert_type=%d, busy=%d\n", lock->ml.type, lock->ml.convert_type, res->state & DLM_LOCK_RES_IN_PROGRESS); @@ -287,6 +294,19 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, status = DLM_DENIED; goto bail; } + + if (lock->ml.type == type && lock->ml.convert_type == LKM_IVMODE) { + mlog(0, "last convert request returned DLM_RECOVERING, but " + "owner has already queued and sent ast to me. res %.*s, " + "(cookie=%u:%llu, type=%d, conv=%d)\n", + res->lockname.len, res->lockname.name, + dlm_get_lock_cookie_node(be64_to_cpu(lock->ml.cookie)), + dlm_get_lock_cookie_seq(be64_to_cpu(lock->ml.cookie)), + lock->ml.type, lock->ml.convert_type); + status = DLM_NORMAL; + goto bail; + } + res->state |= DLM_LOCK_RES_IN_PROGRESS; /* move lock to local convert queue */ /* do not alter lock refcount. switching lists. */ @@ -316,11 +336,19 @@ enum dlm_status dlmconvert_remote(struct dlm_ctxt *dlm, spin_lock(&res->spinlock); res->state &= ~DLM_LOCK_RES_IN_PROGRESS; lock->convert_pending = 0; - /* if it failed, move it back to granted queue */ + /* if it failed, move it back to granted queue. + * if master returns DLM_NORMAL and then down before sending ast, + * it may have already been moved to granted queue, reset to + * DLM_RECOVERING and retry convert */ if (status != DLM_NORMAL) { if (status != DLM_NOTQUEUED) dlm_error(status); dlm_revert_pending_convert(res, lock); + } else if ((res->state & DLM_LOCK_RES_RECOVERING) || + (old_owner != res->owner)) { + mlog(0, "res %.*s is in recovering or has been recovered.\n", + res->lockname.len, res->lockname.name); + status = DLM_RECOVERING; } bail: spin_unlock(&res->spinlock); diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index cd38488a10fc..f6b313898763 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -2083,7 +2083,6 @@ void dlm_move_lockres_to_recovery_list(struct dlm_ctxt *dlm, dlm_lock_get(lock); if (lock->convert_pending) { /* move converting lock back to granted */ - BUG_ON(i != DLM_CONVERTING_LIST); mlog(0, "node died with convert pending " "on %.*s. move back to granted list.\n", res->lockname.len, res->lockname.name); diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 7cb38fdca229..c18ab45f8d21 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1381,44 +1381,6 @@ out: return ret; } -/* - * Will look for holes and unwritten extents in the range starting at - * pos for count bytes (inclusive). - */ -static int ocfs2_check_range_for_holes(struct inode *inode, loff_t pos, - size_t count) -{ - int ret = 0; - unsigned int extent_flags; - u32 cpos, clusters, extent_len, phys_cpos; - struct super_block *sb = inode->i_sb; - - cpos = pos >> OCFS2_SB(sb)->s_clustersize_bits; - clusters = ocfs2_clusters_for_bytes(sb, pos + count) - cpos; - - while (clusters) { - ret = ocfs2_get_clusters(inode, cpos, &phys_cpos, &extent_len, - &extent_flags); - if (ret < 0) { - mlog_errno(ret); - goto out; - } - - if (phys_cpos == 0 || (extent_flags & OCFS2_EXT_UNWRITTEN)) { - ret = 1; - break; - } - - if (extent_len > clusters) - extent_len = clusters; - - clusters -= extent_len; - cpos += extent_len; - } -out: - return ret; -} - static int ocfs2_write_remove_suid(struct inode *inode) { int ret; @@ -2129,18 +2091,12 @@ out: static int ocfs2_prepare_inode_for_write(struct file *file, loff_t pos, - size_t count, - int appending, - int *direct_io, - int *has_refcount) + size_t count) { int ret = 0, meta_level = 0; struct dentry *dentry = file->f_path.dentry; struct inode *inode = d_inode(dentry); loff_t end; - struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); - int full_coherency = !(osb->s_mount_opt & - OCFS2_MOUNT_COHERENCY_BUFFERED); /* * We start with a read level meta lock and only jump to an ex @@ -2189,10 +2145,6 @@ static int ocfs2_prepare_inode_for_write(struct file *file, pos, count, &meta_level); - if (has_refcount) - *has_refcount = 1; - if (direct_io) - *direct_io = 0; } if (ret < 0) { @@ -2200,67 +2152,12 @@ static int ocfs2_prepare_inode_for_write(struct file *file, goto out_unlock; } - /* - * Skip the O_DIRECT checks if we don't need - * them. - */ - if (!direct_io || !(*direct_io)) - break; - - /* - * There's no sane way to do direct writes to an inode - * with inline data. - */ - if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) { - *direct_io = 0; - break; - } - - /* - * Allowing concurrent direct writes means - * i_size changes wouldn't be synchronized, so - * one node could wind up truncating another - * nodes writes. - */ - if (end > i_size_read(inode) && !full_coherency) { - *direct_io = 0; - break; - } - - /* - * Fallback to old way if the feature bit is not set. - */ - if (end > i_size_read(inode) && - !ocfs2_supports_append_dio(osb)) { - *direct_io = 0; - break; - } - - /* - * We don't fill holes during direct io, so - * check for them here. If any are found, the - * caller will have to retake some cluster - * locks and initiate the io as buffered. - */ - ret = ocfs2_check_range_for_holes(inode, pos, count); - if (ret == 1) { - /* - * Fallback to old way if the feature bit is not set. - * Otherwise try dio first and then complete the rest - * request through buffer io. - */ - if (!ocfs2_supports_append_dio(osb)) - *direct_io = 0; - ret = 0; - } else if (ret < 0) - mlog_errno(ret); break; } out_unlock: trace_ocfs2_prepare_inode_for_write(OCFS2_I(inode)->ip_blkno, - pos, appending, count, - direct_io, has_refcount); + pos, count); if (meta_level >= 0) ocfs2_inode_unlock(inode, meta_level); @@ -2272,18 +2169,16 @@ out: static ssize_t ocfs2_file_write_iter(struct kiocb *iocb, struct iov_iter *from) { - int direct_io, appending, rw_level; - int can_do_direct, has_refcount = 0; + int direct_io, rw_level; ssize_t written = 0; ssize_t ret; - size_t count = iov_iter_count(from), orig_count; + size_t count = iov_iter_count(from); struct file *file = iocb->ki_filp; struct inode *inode = file_inode(file); struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); int full_coherency = !(osb->s_mount_opt & OCFS2_MOUNT_COHERENCY_BUFFERED); - int unaligned_dio = 0; - int dropped_dio = 0; + void *saved_ki_complete = NULL; int append_write = ((iocb->ki_pos + count) >= i_size_read(inode) ? 1 : 0); @@ -2296,12 +2191,10 @@ static ssize_t ocfs2_file_write_iter(struct kiocb *iocb, if (count == 0) return 0; - appending = iocb->ki_flags & IOCB_APPEND ? 1 : 0; direct_io = iocb->ki_flags & IOCB_DIRECT ? 1 : 0; inode_lock(inode); -relock: /* * Concurrent O_DIRECT writes are allowed with * mount_option "coherency=buffered". @@ -2334,7 +2227,6 @@ relock: ocfs2_inode_unlock(inode, 1); } - orig_count = iov_iter_count(from); ret = generic_write_checks(iocb, from); if (ret <= 0) { if (ret) @@ -2343,41 +2235,18 @@ relock: } count = ret; - can_do_direct = direct_io; - ret = ocfs2_prepare_inode_for_write(file, iocb->ki_pos, count, appending, - &can_do_direct, &has_refcount); + ret = ocfs2_prepare_inode_for_write(file, iocb->ki_pos, count); if (ret < 0) { mlog_errno(ret); goto out; } - if (direct_io && !is_sync_kiocb(iocb)) - unaligned_dio = ocfs2_is_io_unaligned(inode, count, iocb->ki_pos); - - /* - * We can't complete the direct I/O as requested, fall back to - * buffered I/O. - */ - if (direct_io && !can_do_direct) { - ocfs2_rw_unlock(inode, rw_level); - - rw_level = -1; - - direct_io = 0; - iocb->ki_flags &= ~IOCB_DIRECT; - iov_iter_reexpand(from, orig_count); - dropped_dio = 1; - goto relock; - } - - if (unaligned_dio) { + if (direct_io && !is_sync_kiocb(iocb) && + ocfs2_is_io_unaligned(inode, count, iocb->ki_pos)) { /* - * Wait on previous unaligned aio to complete before - * proceeding. + * Make it a sync io if it's an unaligned aio. */ - mutex_lock(&OCFS2_I(inode)->ip_unaligned_aio); - /* Mark the iocb as needing an unlock in ocfs2_dio_end_io */ - ocfs2_iocb_set_unaligned_aio(iocb); + saved_ki_complete = xchg(&iocb->ki_complete, NULL); } /* communicate with ocfs2_dio_end_io */ @@ -2398,14 +2267,13 @@ relock: */ if ((written == -EIOCBQUEUED) || (!ocfs2_iocb_is_rw_locked(iocb))) { rw_level = -1; - unaligned_dio = 0; } if (unlikely(written <= 0)) - goto no_sync; + goto out; if (((file->f_flags & O_DSYNC) && !direct_io) || - IS_SYNC(inode) || dropped_dio) { + IS_SYNC(inode)) { ret = filemap_fdatawrite_range(file->f_mapping, iocb->ki_pos - written, iocb->ki_pos - 1); @@ -2424,13 +2292,10 @@ relock: iocb->ki_pos - 1); } -no_sync: - if (unaligned_dio && ocfs2_iocb_is_unaligned_aio(iocb)) { - ocfs2_iocb_clear_unaligned_aio(iocb); - mutex_unlock(&OCFS2_I(inode)->ip_unaligned_aio); - } - out: + if (saved_ki_complete) + xchg(&iocb->ki_complete, saved_ki_complete); + if (rw_level != -1) ocfs2_rw_unlock(inode, rw_level); diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index ba495beff1c2..12f4a9e9800f 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -1170,6 +1170,9 @@ static void ocfs2_clear_inode(struct inode *inode) mlog_bug_on_msg(!list_empty(&oi->ip_io_markers), "Clear inode of %llu, inode has io markers\n", (unsigned long long)oi->ip_blkno); + mlog_bug_on_msg(!list_empty(&oi->ip_unwritten_list), + "Clear inode of %llu, inode has unwritten extents\n", + (unsigned long long)oi->ip_blkno); ocfs2_extent_map_trunc(inode, 0); diff --git a/fs/ocfs2/inode.h b/fs/ocfs2/inode.h index 01635e016b3e..d8f3fc8d2551 100644 --- a/fs/ocfs2/inode.h +++ b/fs/ocfs2/inode.h @@ -43,9 +43,6 @@ struct ocfs2_inode_info /* protects extended attribute changes on this inode */ struct rw_semaphore ip_xattr_sem; - /* Number of outstanding AIO's which are not page aligned */ - struct mutex ip_unaligned_aio; - /* These fields are protected by ip_lock */ spinlock_t ip_lock; u32 ip_open_count; @@ -57,6 +54,9 @@ struct ocfs2_inode_info u32 ip_flags; /* see below */ u32 ip_attr; /* inode attributes */ + /* Record unwritten extents during direct io. */ + struct list_head ip_unwritten_list; + /* protected by recovery_lock. */ struct inode *ip_next_orphan; diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index 61b833b721d8..e607419cdfa4 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -231,7 +231,7 @@ void ocfs2_recovery_exit(struct ocfs2_super *osb) /* At this point, we know that no more recovery threads can be * launched, so wait for any recovery completion work to * complete. */ - flush_workqueue(ocfs2_wq); + flush_workqueue(osb->ocfs2_wq); /* * Now that recovery is shut down, and the osb is about to be @@ -1326,7 +1326,7 @@ static void ocfs2_queue_recovery_completion(struct ocfs2_journal *journal, spin_lock(&journal->j_lock); list_add_tail(&item->lri_list, &journal->j_la_cleanups); - queue_work(ocfs2_wq, &journal->j_recovery_work); + queue_work(journal->j_osb->ocfs2_wq, &journal->j_recovery_work); spin_unlock(&journal->j_lock); } @@ -1968,7 +1968,7 @@ static void ocfs2_orphan_scan_work(struct work_struct *work) mutex_lock(&os->os_lock); ocfs2_queue_orphan_scan(osb); if (atomic_read(&os->os_state) == ORPHAN_SCAN_ACTIVE) - queue_delayed_work(ocfs2_wq, &os->os_orphan_scan_work, + queue_delayed_work(osb->ocfs2_wq, &os->os_orphan_scan_work, ocfs2_orphan_scan_timeout()); mutex_unlock(&os->os_lock); } @@ -2008,7 +2008,7 @@ void ocfs2_orphan_scan_start(struct ocfs2_super *osb) atomic_set(&os->os_state, ORPHAN_SCAN_INACTIVE); else { atomic_set(&os->os_state, ORPHAN_SCAN_ACTIVE); - queue_delayed_work(ocfs2_wq, &os->os_orphan_scan_work, + queue_delayed_work(osb->ocfs2_wq, &os->os_orphan_scan_work, ocfs2_orphan_scan_timeout()); } } diff --git a/fs/ocfs2/localalloc.c b/fs/ocfs2/localalloc.c index 7d62c43a2c3e..fe0d1f9571bb 100644 --- a/fs/ocfs2/localalloc.c +++ b/fs/ocfs2/localalloc.c @@ -386,7 +386,7 @@ void ocfs2_shutdown_local_alloc(struct ocfs2_super *osb) struct ocfs2_dinode *alloc = NULL; cancel_delayed_work(&osb->la_enable_wq); - flush_workqueue(ocfs2_wq); + flush_workqueue(osb->ocfs2_wq); if (osb->local_alloc_state == OCFS2_LA_UNUSED) goto out; @@ -1085,7 +1085,7 @@ static int ocfs2_recalc_la_window(struct ocfs2_super *osb, } else { osb->local_alloc_state = OCFS2_LA_DISABLED; } - queue_delayed_work(ocfs2_wq, &osb->la_enable_wq, + queue_delayed_work(osb->ocfs2_wq, &osb->la_enable_wq, OCFS2_LA_ENABLE_INTERVAL); goto out_unlock; } diff --git a/fs/ocfs2/mmap.c b/fs/ocfs2/mmap.c index 77ebc2bc1cca..9ea081f4e6e4 100644 --- a/fs/ocfs2/mmap.c +++ b/fs/ocfs2/mmap.c @@ -104,8 +104,8 @@ static int __ocfs2_page_mkwrite(struct file *file, struct buffer_head *di_bh, if (page->index == last_index) len = ((size - 1) & ~PAGE_CACHE_MASK) + 1; - ret = ocfs2_write_begin_nolock(file, mapping, pos, len, 0, &locked_page, - &fsdata, di_bh, page); + ret = ocfs2_write_begin_nolock(mapping, pos, len, OCFS2_WRITE_MMAP, + &locked_page, &fsdata, di_bh, page); if (ret) { if (ret != -ENOSPC) mlog_errno(ret); diff --git a/fs/ocfs2/ocfs2.h b/fs/ocfs2/ocfs2.h index 7a0126267847..6cf6538a0651 100644 --- a/fs/ocfs2/ocfs2.h +++ b/fs/ocfs2/ocfs2.h @@ -464,6 +464,14 @@ struct ocfs2_super struct ocfs2_refcount_tree *osb_ref_tree_lru; struct mutex system_file_mutex; + + /* + * OCFS2 needs to schedule several different types of work which + * require cluster locking, disk I/O, recovery waits, etc. Since these + * types of work tend to be heavy we avoid using the kernel events + * workqueue and schedule on our own. + */ + struct workqueue_struct *ocfs2_wq; }; #define OCFS2_SB(sb) ((struct ocfs2_super *)(sb)->s_fs_info) diff --git a/fs/ocfs2/ocfs2_trace.h b/fs/ocfs2/ocfs2_trace.h index 24b7e7f591dc..f8f5fc5e6c05 100644 --- a/fs/ocfs2/ocfs2_trace.h +++ b/fs/ocfs2/ocfs2_trace.h @@ -1450,28 +1450,20 @@ DEFINE_OCFS2_ULL_ULL_ULL_EVENT(ocfs2_remove_inode_range); TRACE_EVENT(ocfs2_prepare_inode_for_write, TP_PROTO(unsigned long long ino, unsigned long long saved_pos, - int appending, unsigned long count, - int *direct_io, int *has_refcount), - TP_ARGS(ino, saved_pos, appending, count, direct_io, has_refcount), + unsigned long count), + TP_ARGS(ino, saved_pos, count), TP_STRUCT__entry( __field(unsigned long long, ino) __field(unsigned long long, saved_pos) - __field(int, appending) __field(unsigned long, count) - __field(int, direct_io) - __field(int, has_refcount) ), TP_fast_assign( __entry->ino = ino; __entry->saved_pos = saved_pos; - __entry->appending = appending; __entry->count = count; - __entry->direct_io = direct_io ? *direct_io : -1; - __entry->has_refcount = has_refcount ? *has_refcount : -1; ), - TP_printk("%llu %llu %d %lu %d %d", __entry->ino, - __entry->saved_pos, __entry->appending, __entry->count, - __entry->direct_io, __entry->has_refcount) + TP_printk("%llu %llu %lu", __entry->ino, + __entry->saved_pos, __entry->count) ); DEFINE_OCFS2_INT_EVENT(generic_file_aio_read_ret); diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 91bc674203ed..3892f3c079ca 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -726,7 +726,7 @@ static int ocfs2_release_dquot(struct dquot *dquot) dqgrab(dquot); /* First entry on list -> queue work */ if (llist_add(&OCFS2_DQUOT(dquot)->list, &osb->dquot_drop_list)) - queue_work(ocfs2_wq, &osb->dquot_drop_work); + queue_work(osb->ocfs2_wq, &osb->dquot_drop_work); goto out; } status = ocfs2_lock_global_qf(oinfo, 1); diff --git a/fs/ocfs2/resize.c b/fs/ocfs2/resize.c index 576b9a04873f..18451e0fab81 100644 --- a/fs/ocfs2/resize.c +++ b/fs/ocfs2/resize.c @@ -196,7 +196,7 @@ static int update_backups(struct inode * inode, u32 clusters, char *data) for (i = 0; i < OCFS2_MAX_BACKUP_SUPERBLOCKS; i++) { blkno = ocfs2_backup_super_blkno(inode->i_sb, i); cluster = ocfs2_blocks_to_clusters(inode->i_sb, blkno); - if (cluster > clusters) + if (cluster >= clusters) break; ret = ocfs2_read_blocks_sync(osb, blkno, 1, &backup); diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index ccc9386c42c5..7db631e1c8b0 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -80,12 +80,6 @@ static struct kmem_cache *ocfs2_inode_cachep; struct kmem_cache *ocfs2_dquot_cachep; struct kmem_cache *ocfs2_qf_chunk_cachep; -/* OCFS2 needs to schedule several different types of work which - * require cluster locking, disk I/O, recovery waits, etc. Since these - * types of work tend to be heavy we avoid using the kernel events - * workqueue and schedule on our own. */ -struct workqueue_struct *ocfs2_wq = NULL; - static struct dentry *ocfs2_debugfs_root; MODULE_AUTHOR("Oracle"); @@ -1613,33 +1607,25 @@ static int __init ocfs2_init(void) if (status < 0) goto out2; - ocfs2_wq = create_singlethread_workqueue("ocfs2_wq"); - if (!ocfs2_wq) { - status = -ENOMEM; - goto out3; - } - ocfs2_debugfs_root = debugfs_create_dir("ocfs2", NULL); if (!ocfs2_debugfs_root) { status = -ENOMEM; mlog(ML_ERROR, "Unable to create ocfs2 debugfs root.\n"); - goto out4; + goto out3; } ocfs2_set_locking_protocol(); status = register_quota_format(&ocfs2_quota_format); if (status < 0) - goto out4; + goto out3; status = register_filesystem(&ocfs2_fs_type); if (!status) return 0; unregister_quota_format(&ocfs2_quota_format); -out4: - destroy_workqueue(ocfs2_wq); - debugfs_remove(ocfs2_debugfs_root); out3: + debugfs_remove(ocfs2_debugfs_root); ocfs2_free_mem_caches(); out2: exit_ocfs2_uptodate_cache(); @@ -1650,11 +1636,6 @@ out1: static void __exit ocfs2_exit(void) { - if (ocfs2_wq) { - flush_workqueue(ocfs2_wq); - destroy_workqueue(ocfs2_wq); - } - unregister_quota_format(&ocfs2_quota_format); debugfs_remove(ocfs2_debugfs_root); @@ -1745,8 +1726,8 @@ static void ocfs2_inode_init_once(void *data) spin_lock_init(&oi->ip_lock); ocfs2_extent_map_init(&oi->vfs_inode); INIT_LIST_HEAD(&oi->ip_io_markers); + INIT_LIST_HEAD(&oi->ip_unwritten_list); oi->ip_dir_start_lookup = 0; - mutex_init(&oi->ip_unaligned_aio); init_rwsem(&oi->ip_alloc_sem); init_rwsem(&oi->ip_xattr_sem); mutex_init(&oi->ip_io_mutex); @@ -2349,6 +2330,12 @@ static int ocfs2_initialize_super(struct super_block *sb, } cleancache_init_shared_fs(sb); + osb->ocfs2_wq = create_singlethread_workqueue("ocfs2_wq"); + if (!osb->ocfs2_wq) { + status = -ENOMEM; + mlog_errno(status); + } + bail: return status; } @@ -2536,6 +2523,12 @@ static void ocfs2_delete_osb(struct ocfs2_super *osb) { /* This function assumes that the caller has the main osb resource */ + /* ocfs2_initializer_super have already created this workqueue */ + if (osb->ocfs2_wq) { + flush_workqueue(osb->ocfs2_wq); + destroy_workqueue(osb->ocfs2_wq); + } + ocfs2_free_slot_info(osb); kfree(osb->osb_orphan_wipes); diff --git a/fs/ocfs2/super.h b/fs/ocfs2/super.h index b477d0b1c7b6..b023e4f3d740 100644 --- a/fs/ocfs2/super.h +++ b/fs/ocfs2/super.h @@ -26,8 +26,6 @@ #ifndef OCFS2_SUPER_H #define OCFS2_SUPER_H -extern struct workqueue_struct *ocfs2_wq; - int ocfs2_publish_get_mount_state(struct ocfs2_super *osb, int node_num); diff --git a/fs/orangefs/Kconfig b/fs/orangefs/Kconfig new file mode 100644 index 000000000000..1554c02489de --- /dev/null +++ b/fs/orangefs/Kconfig @@ -0,0 +1,6 @@ +config ORANGEFS_FS + tristate "ORANGEFS (Powered by PVFS) support" + select FS_POSIX_ACL + help + Orange is a parallel file system designed for use on high end + computing (HEC) systems. diff --git a/fs/orangefs/Makefile b/fs/orangefs/Makefile new file mode 100644 index 000000000000..a9d6a968fe6d --- /dev/null +++ b/fs/orangefs/Makefile @@ -0,0 +1,10 @@ +# +# Makefile for the ORANGEFS filesystem. +# + +obj-$(CONFIG_ORANGEFS_FS) += orangefs.o + +orangefs-objs := acl.o file.o orangefs-cache.o orangefs-utils.o xattr.o \ + dcache.o inode.o orangefs-sysfs.o orangefs-mod.o super.o \ + devorangefs-req.o namei.o symlink.o dir.o orangefs-bufmap.o \ + orangefs-debugfs.o waitqueue.o diff --git a/fs/orangefs/acl.c b/fs/orangefs/acl.c new file mode 100644 index 000000000000..03f89dbb2512 --- /dev/null +++ b/fs/orangefs/acl.c @@ -0,0 +1,175 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" +#include <linux/posix_acl_xattr.h> +#include <linux/fs_struct.h> + +struct posix_acl *orangefs_get_acl(struct inode *inode, int type) +{ + struct posix_acl *acl; + int ret; + char *key = NULL, *value = NULL; + + switch (type) { + case ACL_TYPE_ACCESS: + key = ORANGEFS_XATTR_NAME_ACL_ACCESS; + break; + case ACL_TYPE_DEFAULT: + key = ORANGEFS_XATTR_NAME_ACL_DEFAULT; + break; + default: + gossip_err("orangefs_get_acl: bogus value of type %d\n", type); + return ERR_PTR(-EINVAL); + } + /* + * Rather than incurring a network call just to determine the exact + * length of the attribute, I just allocate a max length to save on + * the network call. Conceivably, we could pass NULL to + * orangefs_inode_getxattr() to probe the length of the value, but + * I don't do that for now. + */ + value = kmalloc(ORANGEFS_MAX_XATTR_VALUELEN, GFP_KERNEL); + if (value == NULL) + return ERR_PTR(-ENOMEM); + + gossip_debug(GOSSIP_ACL_DEBUG, + "inode %pU, key %s, type %d\n", + get_khandle_from_ino(inode), + key, + type); + ret = orangefs_inode_getxattr(inode, + "", + key, + value, + ORANGEFS_MAX_XATTR_VALUELEN); + /* if the key exists, convert it to an in-memory rep */ + if (ret > 0) { + acl = posix_acl_from_xattr(&init_user_ns, value, ret); + } else if (ret == -ENODATA || ret == -ENOSYS) { + acl = NULL; + } else { + gossip_err("inode %pU retrieving acl's failed with error %d\n", + get_khandle_from_ino(inode), + ret); + acl = ERR_PTR(ret); + } + /* kfree(NULL) is safe, so don't worry if value ever got used */ + kfree(value); + return acl; +} + +int orangefs_set_acl(struct inode *inode, struct posix_acl *acl, int type) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + int error = 0; + void *value = NULL; + size_t size = 0; + const char *name = NULL; + + switch (type) { + case ACL_TYPE_ACCESS: + name = ORANGEFS_XATTR_NAME_ACL_ACCESS; + if (acl) { + umode_t mode = inode->i_mode; + /* + * can we represent this with the traditional file + * mode permission bits? + */ + error = posix_acl_equiv_mode(acl, &mode); + if (error < 0) { + gossip_err("%s: posix_acl_equiv_mode err: %d\n", + __func__, + error); + return error; + } + + if (inode->i_mode != mode) + SetModeFlag(orangefs_inode); + inode->i_mode = mode; + mark_inode_dirty_sync(inode); + if (error == 0) + acl = NULL; + } + break; + case ACL_TYPE_DEFAULT: + name = ORANGEFS_XATTR_NAME_ACL_DEFAULT; + break; + default: + gossip_err("%s: invalid type %d!\n", __func__, type); + return -EINVAL; + } + + gossip_debug(GOSSIP_ACL_DEBUG, + "%s: inode %pU, key %s type %d\n", + __func__, get_khandle_from_ino(inode), + name, + type); + + if (acl) { + size = posix_acl_xattr_size(acl->a_count); + value = kmalloc(size, GFP_KERNEL); + if (!value) + return -ENOMEM; + + error = posix_acl_to_xattr(&init_user_ns, acl, value, size); + if (error < 0) + goto out; + } + + gossip_debug(GOSSIP_ACL_DEBUG, + "%s: name %s, value %p, size %zd, acl %p\n", + __func__, name, value, size, acl); + /* + * Go ahead and set the extended attribute now. NOTE: Suppose acl + * was NULL, then value will be NULL and size will be 0 and that + * will xlate to a removexattr. However, we don't want removexattr + * complain if attributes does not exist. + */ + error = orangefs_inode_setxattr(inode, "", name, value, size, 0); + +out: + kfree(value); + if (!error) + set_cached_acl(inode, type, acl); + return error; +} + +int orangefs_init_acl(struct inode *inode, struct inode *dir) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct posix_acl *default_acl, *acl; + umode_t mode = inode->i_mode; + int error = 0; + + ClearModeFlag(orangefs_inode); + + error = posix_acl_create(dir, &mode, &default_acl, &acl); + if (error) + return error; + + if (default_acl) { + error = orangefs_set_acl(inode, default_acl, ACL_TYPE_DEFAULT); + posix_acl_release(default_acl); + } + + if (acl) { + if (!error) + error = orangefs_set_acl(inode, acl, ACL_TYPE_ACCESS); + posix_acl_release(acl); + } + + /* If mode of the inode was changed, then do a forcible ->setattr */ + if (mode != inode->i_mode) { + SetModeFlag(orangefs_inode); + inode->i_mode = mode; + orangefs_flush_inode(inode); + } + + return error; +} diff --git a/fs/orangefs/dcache.c b/fs/orangefs/dcache.c new file mode 100644 index 000000000000..5dfc4f3cfe68 --- /dev/null +++ b/fs/orangefs/dcache.c @@ -0,0 +1,138 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Implementation of dentry (directory cache) functions. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" + +/* Returns 1 if dentry can still be trusted, else 0. */ +static int orangefs_revalidate_lookup(struct dentry *dentry) +{ + struct dentry *parent_dentry = dget_parent(dentry); + struct inode *parent_inode = parent_dentry->d_inode; + struct orangefs_inode_s *parent = ORANGEFS_I(parent_inode); + struct inode *inode = dentry->d_inode; + struct orangefs_kernel_op_s *new_op; + int ret = 0; + int err = 0; + + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s: attempting lookup.\n", __func__); + + new_op = op_alloc(ORANGEFS_VFS_OP_LOOKUP); + if (!new_op) + goto out_put_parent; + + new_op->upcall.req.lookup.sym_follow = ORANGEFS_LOOKUP_LINK_NO_FOLLOW; + new_op->upcall.req.lookup.parent_refn = parent->refn; + strncpy(new_op->upcall.req.lookup.d_name, + dentry->d_name.name, + ORANGEFS_NAME_MAX); + + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d interrupt flag [%d]\n", + __FILE__, + __func__, + __LINE__, + get_interruptible_flag(parent_inode)); + + err = service_operation(new_op, "orangefs_lookup", + get_interruptible_flag(parent_inode)); + + /* Positive dentry: reject if error or not the same inode. */ + if (inode) { + if (err) { + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d lookup failure.\n", + __FILE__, __func__, __LINE__); + goto out_drop; + } + if (!match_handle(new_op->downcall.resp.lookup.refn.khandle, + inode)) { + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d no match.\n", + __FILE__, __func__, __LINE__); + goto out_drop; + } + + /* Negative dentry: reject if success or error other than ENOENT. */ + } else { + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s: negative dentry.\n", + __func__); + if (!err || err != -ENOENT) { + if (new_op->downcall.status != 0) + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s:%s:%d lookup failure.\n", + __FILE__, __func__, __LINE__); + goto out_drop; + } + } + + ret = 1; +out_release_op: + op_release(new_op); +out_put_parent: + dput(parent_dentry); + return ret; +out_drop: + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s:%s:%d revalidate failed\n", + __FILE__, __func__, __LINE__); + goto out_release_op; +} + +/* + * Verify that dentry is valid. + * + * Should return 1 if dentry can still be trusted, else 0. + */ +static int orangefs_d_revalidate(struct dentry *dentry, unsigned int flags) +{ + int ret; + + if (flags & LOOKUP_RCU) + return -ECHILD; + + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s: called on dentry %p.\n", + __func__, dentry); + + /* skip root handle lookups. */ + if (dentry->d_inode && is_root_handle(dentry->d_inode)) + return 1; + + /* + * If this passes, the positive dentry still exists or the negative + * dentry still does not exist. + */ + if (!orangefs_revalidate_lookup(dentry)) + return 0; + + /* We do not need to continue with negative dentries. */ + if (!dentry->d_inode) + goto out; + + /* Now we must perform a getattr to validate the inode contents. */ + + ret = orangefs_inode_check_changed(dentry->d_inode); + if (ret < 0) { + gossip_debug(GOSSIP_DCACHE_DEBUG, "%s:%s:%d getattr failure.\n", + __FILE__, __func__, __LINE__); + return 0; + } + if (ret == 0) + return 0; + +out: + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s: negative dentry or positive dentry and inode valid.\n", + __func__); + return 1; +} + +const struct dentry_operations orangefs_dentry_operations = { + .d_revalidate = orangefs_d_revalidate, +}; diff --git a/fs/orangefs/devorangefs-req.c b/fs/orangefs/devorangefs-req.c new file mode 100644 index 000000000000..db170beba797 --- /dev/null +++ b/fs/orangefs/devorangefs-req.c @@ -0,0 +1,943 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * Changes by Acxiom Corporation to add protocol version to kernel + * communication, Copyright Acxiom Corporation, 2005. + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-dev-proto.h" +#include "orangefs-bufmap.h" + +#include <linux/debugfs.h> +#include <linux/slab.h> + +/* this file implements the /dev/pvfs2-req device node */ + +static int open_access_count; + +#define DUMP_DEVICE_ERROR() \ +do { \ + gossip_err("*****************************************************\n");\ + gossip_err("ORANGEFS Device Error: You cannot open the device file "); \ + gossip_err("\n/dev/%s more than once. Please make sure that\nthere " \ + "are no ", ORANGEFS_REQDEVICE_NAME); \ + gossip_err("instances of a program using this device\ncurrently " \ + "running. (You must verify this!)\n"); \ + gossip_err("For example, you can use the lsof program as follows:\n");\ + gossip_err("'lsof | grep %s' (run this as root)\n", \ + ORANGEFS_REQDEVICE_NAME); \ + gossip_err(" open_access_count = %d\n", open_access_count); \ + gossip_err("*****************************************************\n");\ +} while (0) + +static int hash_func(__u64 tag, int table_size) +{ + return do_div(tag, (unsigned int)table_size); +} + +static void orangefs_devreq_add_op(struct orangefs_kernel_op_s *op) +{ + int index = hash_func(op->tag, hash_table_size); + + list_add_tail(&op->list, &htable_ops_in_progress[index]); +} + +/* + * find the op with this tag and remove it from the in progress + * hash table. + */ +static struct orangefs_kernel_op_s *orangefs_devreq_remove_op(__u64 tag) +{ + struct orangefs_kernel_op_s *op, *next; + int index; + + index = hash_func(tag, hash_table_size); + + spin_lock(&htable_ops_in_progress_lock); + list_for_each_entry_safe(op, + next, + &htable_ops_in_progress[index], + list) { + if (op->tag == tag && !op_state_purged(op) && + !op_state_given_up(op)) { + list_del_init(&op->list); + spin_unlock(&htable_ops_in_progress_lock); + return op; + } + } + + spin_unlock(&htable_ops_in_progress_lock); + return NULL; +} + +/* Returns whether any FS are still pending remounted */ +static int mark_all_pending_mounts(void) +{ + int unmounted = 1; + struct orangefs_sb_info_s *orangefs_sb = NULL; + + spin_lock(&orangefs_superblocks_lock); + list_for_each_entry(orangefs_sb, &orangefs_superblocks, list) { + /* All of these file system require a remount */ + orangefs_sb->mount_pending = 1; + unmounted = 0; + } + spin_unlock(&orangefs_superblocks_lock); + return unmounted; +} + +/* + * Determine if a given file system needs to be remounted or not + * Returns -1 on error + * 0 if already mounted + * 1 if needs remount + */ +static int fs_mount_pending(__s32 fsid) +{ + int mount_pending = -1; + struct orangefs_sb_info_s *orangefs_sb = NULL; + + spin_lock(&orangefs_superblocks_lock); + list_for_each_entry(orangefs_sb, &orangefs_superblocks, list) { + if (orangefs_sb->fs_id == fsid) { + mount_pending = orangefs_sb->mount_pending; + break; + } + } + spin_unlock(&orangefs_superblocks_lock); + return mount_pending; +} + +static int orangefs_devreq_open(struct inode *inode, struct file *file) +{ + int ret = -EINVAL; + + if (!(file->f_flags & O_NONBLOCK)) { + gossip_err("%s: device cannot be opened in blocking mode\n", + __func__); + goto out; + } + ret = -EACCES; + gossip_debug(GOSSIP_DEV_DEBUG, "client-core: opening device\n"); + mutex_lock(&devreq_mutex); + + if (open_access_count == 0) { + open_access_count = 1; + ret = 0; + } else { + DUMP_DEVICE_ERROR(); + } + mutex_unlock(&devreq_mutex); + +out: + + gossip_debug(GOSSIP_DEV_DEBUG, + "pvfs2-client-core: open device complete (ret = %d)\n", + ret); + return ret; +} + +/* Function for read() callers into the device */ +static ssize_t orangefs_devreq_read(struct file *file, + char __user *buf, + size_t count, loff_t *offset) +{ + struct orangefs_kernel_op_s *op, *temp; + __s32 proto_ver = ORANGEFS_KERNEL_PROTO_VERSION; + static __s32 magic = ORANGEFS_DEVREQ_MAGIC; + struct orangefs_kernel_op_s *cur_op = NULL; + unsigned long ret; + + /* We do not support blocking IO. */ + if (!(file->f_flags & O_NONBLOCK)) { + gossip_err("%s: blocking read from client-core.\n", + __func__); + return -EINVAL; + } + + /* + * The client will do an ioctl to find MAX_DEV_REQ_UPSIZE, then + * always read with that size buffer. + */ + if (count != MAX_DEV_REQ_UPSIZE) { + gossip_err("orangefs: client-core tried to read wrong size\n"); + return -EINVAL; + } + +restart: + /* Get next op (if any) from top of list. */ + spin_lock(&orangefs_request_list_lock); + list_for_each_entry_safe(op, temp, &orangefs_request_list, list) { + __s32 fsid; + /* This lock is held past the end of the loop when we break. */ + spin_lock(&op->lock); + if (unlikely(op_state_purged(op) || op_state_given_up(op))) { + spin_unlock(&op->lock); + continue; + } + + fsid = fsid_of_op(op); + if (fsid != ORANGEFS_FS_ID_NULL) { + int ret; + /* Skip ops whose filesystem needs to be mounted. */ + ret = fs_mount_pending(fsid); + if (ret == 1) { + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: mount pending, skipping op tag " + "%llu %s\n", + __func__, + llu(op->tag), + get_opname_string(op)); + spin_unlock(&op->lock); + continue; + /* + * Skip ops whose filesystem we don't know about unless + * it is being mounted. + */ + /* XXX: is there a better way to detect this? */ + } else if (ret == -1 && + !(op->upcall.type == + ORANGEFS_VFS_OP_FS_MOUNT || + op->upcall.type == + ORANGEFS_VFS_OP_GETATTR)) { + gossip_debug(GOSSIP_DEV_DEBUG, + "orangefs: skipping op tag %llu %s\n", + llu(op->tag), get_opname_string(op)); + gossip_err( + "orangefs: ERROR: fs_mount_pending %d\n", + fsid); + spin_unlock(&op->lock); + continue; + } + } + /* + * Either this op does not pertain to a filesystem, is mounting + * a filesystem, or pertains to a mounted filesystem. Let it + * through. + */ + cur_op = op; + break; + } + + /* + * At this point we either have a valid op and can continue or have not + * found an op and must ask the client to try again later. + */ + if (!cur_op) { + spin_unlock(&orangefs_request_list_lock); + return -EAGAIN; + } + + gossip_debug(GOSSIP_DEV_DEBUG, "%s: reading op tag %llu %s\n", + __func__, + llu(cur_op->tag), + get_opname_string(cur_op)); + + /* + * Such an op should never be on the list in the first place. If so, we + * will abort. + */ + if (op_state_in_progress(cur_op) || op_state_serviced(cur_op)) { + gossip_err("orangefs: ERROR: Current op already queued.\n"); + list_del_init(&cur_op->list); + spin_unlock(&cur_op->lock); + spin_unlock(&orangefs_request_list_lock); + return -EAGAIN; + } + + list_del_init(&cur_op->list); + spin_unlock(&orangefs_request_list_lock); + + spin_unlock(&cur_op->lock); + + /* Push the upcall out. */ + ret = copy_to_user(buf, &proto_ver, sizeof(__s32)); + if (ret != 0) + goto error; + ret = copy_to_user(buf+sizeof(__s32), &magic, sizeof(__s32)); + if (ret != 0) + goto error; + ret = copy_to_user(buf+2 * sizeof(__s32), &cur_op->tag, sizeof(__u64)); + if (ret != 0) + goto error; + ret = copy_to_user(buf+2*sizeof(__s32)+sizeof(__u64), &cur_op->upcall, + sizeof(struct orangefs_upcall_s)); + if (ret != 0) + goto error; + + spin_lock(&htable_ops_in_progress_lock); + spin_lock(&cur_op->lock); + if (unlikely(op_state_given_up(cur_op))) { + spin_unlock(&cur_op->lock); + spin_unlock(&htable_ops_in_progress_lock); + complete(&cur_op->waitq); + goto restart; + } + + /* + * Set the operation to be in progress and move it between lists since + * it has been sent to the client. + */ + set_op_state_inprogress(cur_op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: 1 op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(cur_op), + cur_op->op_state, + current->comm); + orangefs_devreq_add_op(cur_op); + spin_unlock(&cur_op->lock); + spin_unlock(&htable_ops_in_progress_lock); + + /* The client only asks to read one size buffer. */ + return MAX_DEV_REQ_UPSIZE; +error: + /* + * We were unable to copy the op data to the client. Put the op back in + * list. If client has crashed, the op will be purged later when the + * device is released. + */ + gossip_err("orangefs: Failed to copy data to user space\n"); + spin_lock(&orangefs_request_list_lock); + spin_lock(&cur_op->lock); + if (likely(!op_state_given_up(cur_op))) { + set_op_state_waiting(cur_op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: 2 op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(cur_op), + cur_op->op_state, + current->comm); + list_add(&cur_op->list, &orangefs_request_list); + spin_unlock(&cur_op->lock); + } else { + spin_unlock(&cur_op->lock); + complete(&cur_op->waitq); + } + spin_unlock(&orangefs_request_list_lock); + return -EFAULT; +} + +/* + * Function for writev() callers into the device. + * + * Userspace should have written: + * - __u32 version + * - __u32 magic + * - __u64 tag + * - struct orangefs_downcall_s + * - trailer buffer (in the case of READDIR operations) + */ +static ssize_t orangefs_devreq_write_iter(struct kiocb *iocb, + struct iov_iter *iter) +{ + ssize_t ret; + struct orangefs_kernel_op_s *op = NULL; + struct { + __u32 version; + __u32 magic; + __u64 tag; + } head; + int total = ret = iov_iter_count(iter); + int n; + int downcall_size = sizeof(struct orangefs_downcall_s); + int head_size = sizeof(head); + + gossip_debug(GOSSIP_DEV_DEBUG, "%s: total:%d: ret:%zd:\n", + __func__, + total, + ret); + + if (total < MAX_DEV_REQ_DOWNSIZE) { + gossip_err("%s: total:%d: must be at least:%u:\n", + __func__, + total, + (unsigned int) MAX_DEV_REQ_DOWNSIZE); + return -EFAULT; + } + + n = copy_from_iter(&head, head_size, iter); + if (n < head_size) { + gossip_err("%s: failed to copy head.\n", __func__); + return -EFAULT; + } + + if (head.version < ORANGEFS_MINIMUM_USERSPACE_VERSION) { + gossip_err("%s: userspace claims version" + "%d, minimum version required: %d.\n", + __func__, + head.version, + ORANGEFS_MINIMUM_USERSPACE_VERSION); + return -EPROTO; + } + + if (head.magic != ORANGEFS_DEVREQ_MAGIC) { + gossip_err("Error: Device magic number does not match.\n"); + return -EPROTO; + } + + /* remove the op from the in progress hash table */ + op = orangefs_devreq_remove_op(head.tag); + if (!op) { + gossip_err("WARNING: No one's waiting for tag %llu\n", + llu(head.tag)); + return ret; + } + + n = copy_from_iter(&op->downcall, downcall_size, iter); + if (n != downcall_size) { + gossip_err("%s: failed to copy downcall.\n", __func__); + goto Efault; + } + + if (op->downcall.status) + goto wakeup; + + /* + * We've successfully peeled off the head and the downcall. + * Something has gone awry if total doesn't equal the + * sum of head_size, downcall_size and trailer_size. + */ + if ((head_size + downcall_size + op->downcall.trailer_size) != total) { + gossip_err("%s: funky write, head_size:%d" + ": downcall_size:%d: trailer_size:%lld" + ": total size:%d:\n", + __func__, + head_size, + downcall_size, + op->downcall.trailer_size, + total); + goto Efault; + } + + /* Only READDIR operations should have trailers. */ + if ((op->downcall.type != ORANGEFS_VFS_OP_READDIR) && + (op->downcall.trailer_size != 0)) { + gossip_err("%s: %x operation with trailer.", + __func__, + op->downcall.type); + goto Efault; + } + + /* READDIR operations should always have trailers. */ + if ((op->downcall.type == ORANGEFS_VFS_OP_READDIR) && + (op->downcall.trailer_size == 0)) { + gossip_err("%s: %x operation with no trailer.", + __func__, + op->downcall.type); + goto Efault; + } + + if (op->downcall.type != ORANGEFS_VFS_OP_READDIR) + goto wakeup; + + op->downcall.trailer_buf = + vmalloc(op->downcall.trailer_size); + if (op->downcall.trailer_buf == NULL) { + gossip_err("%s: failed trailer vmalloc.\n", + __func__); + goto Enomem; + } + memset(op->downcall.trailer_buf, 0, op->downcall.trailer_size); + n = copy_from_iter(op->downcall.trailer_buf, + op->downcall.trailer_size, + iter); + if (n != op->downcall.trailer_size) { + gossip_err("%s: failed to copy trailer.\n", __func__); + vfree(op->downcall.trailer_buf); + goto Efault; + } + +wakeup: + /* + * Return to vfs waitqueue, and back to service_operation + * through wait_for_matching_downcall. + */ + spin_lock(&op->lock); + if (unlikely(op_is_cancel(op))) { + spin_unlock(&op->lock); + put_cancel(op); + } else if (unlikely(op_state_given_up(op))) { + spin_unlock(&op->lock); + complete(&op->waitq); + } else { + set_op_state_serviced(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + spin_unlock(&op->lock); + } + return ret; + +Efault: + op->downcall.status = -(ORANGEFS_ERROR_BIT | 9); + ret = -EFAULT; + goto wakeup; + +Enomem: + op->downcall.status = -(ORANGEFS_ERROR_BIT | 8); + ret = -ENOMEM; + goto wakeup; +} + +/* + * NOTE: gets called when the last reference to this device is dropped. + * Using the open_access_count variable, we enforce a reference count + * on this file so that it can be opened by only one process at a time. + * the devreq_mutex is used to make sure all i/o has completed + * before we call orangefs_bufmap_finalize, and similar such tricky + * situations + */ +static int orangefs_devreq_release(struct inode *inode, struct file *file) +{ + int unmounted = 0; + + gossip_debug(GOSSIP_DEV_DEBUG, + "%s:pvfs2-client-core: exiting, closing device\n", + __func__); + + mutex_lock(&devreq_mutex); + orangefs_bufmap_finalize(); + + open_access_count = -1; + + unmounted = mark_all_pending_mounts(); + gossip_debug(GOSSIP_DEV_DEBUG, "ORANGEFS Device Close: Filesystem(s) %s\n", + (unmounted ? "UNMOUNTED" : "MOUNTED")); + + purge_waiting_ops(); + purge_inprogress_ops(); + + orangefs_bufmap_run_down(); + + gossip_debug(GOSSIP_DEV_DEBUG, + "pvfs2-client-core: device close complete\n"); + open_access_count = 0; + mutex_unlock(&devreq_mutex); + return 0; +} + +int is_daemon_in_service(void) +{ + int in_service; + + /* + * What this function does is checks if client-core is alive + * based on the access count we maintain on the device. + */ + mutex_lock(&devreq_mutex); + in_service = open_access_count == 1 ? 0 : -EIO; + mutex_unlock(&devreq_mutex); + return in_service; +} + +bool __is_daemon_in_service(void) +{ + return open_access_count == 1; +} + +static inline long check_ioctl_command(unsigned int command) +{ + /* Check for valid ioctl codes */ + if (_IOC_TYPE(command) != ORANGEFS_DEV_MAGIC) { + gossip_err("device ioctl magic numbers don't match! Did you rebuild pvfs2-client-core/libpvfs2? [cmd %x, magic %x != %x]\n", + command, + _IOC_TYPE(command), + ORANGEFS_DEV_MAGIC); + return -EINVAL; + } + /* and valid ioctl commands */ + if (_IOC_NR(command) >= ORANGEFS_DEV_MAXNR || _IOC_NR(command) <= 0) { + gossip_err("Invalid ioctl command number [%d >= %d]\n", + _IOC_NR(command), ORANGEFS_DEV_MAXNR); + return -ENOIOCTLCMD; + } + return 0; +} + +static long dispatch_ioctl_command(unsigned int command, unsigned long arg) +{ + static __s32 magic = ORANGEFS_DEVREQ_MAGIC; + static __s32 max_up_size = MAX_DEV_REQ_UPSIZE; + static __s32 max_down_size = MAX_DEV_REQ_DOWNSIZE; + struct ORANGEFS_dev_map_desc user_desc; + int ret = 0; + struct dev_mask_info_s mask_info = { 0 }; + struct dev_mask2_info_s mask2_info = { 0, 0 }; + int upstream_kmod = 1; + struct orangefs_sb_info_s *orangefs_sb; + + /* mtmoore: add locking here */ + + switch (command) { + case ORANGEFS_DEV_GET_MAGIC: + return ((put_user(magic, (__s32 __user *) arg) == -EFAULT) ? + -EIO : + 0); + case ORANGEFS_DEV_GET_MAX_UPSIZE: + return ((put_user(max_up_size, + (__s32 __user *) arg) == -EFAULT) ? + -EIO : + 0); + case ORANGEFS_DEV_GET_MAX_DOWNSIZE: + return ((put_user(max_down_size, + (__s32 __user *) arg) == -EFAULT) ? + -EIO : + 0); + case ORANGEFS_DEV_MAP: + ret = copy_from_user(&user_desc, + (struct ORANGEFS_dev_map_desc __user *) + arg, + sizeof(struct ORANGEFS_dev_map_desc)); + /* WTF -EIO and not -EFAULT? */ + return ret ? -EIO : orangefs_bufmap_initialize(&user_desc); + case ORANGEFS_DEV_REMOUNT_ALL: + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: got ORANGEFS_DEV_REMOUNT_ALL\n", + __func__); + + /* + * remount all mounted orangefs volumes to regain the lost + * dynamic mount tables (if any) -- NOTE: this is done + * without keeping the superblock list locked due to the + * upcall/downcall waiting. also, the request mutex is + * used to ensure that no operations will be serviced until + * all of the remounts are serviced (to avoid ops between + * mounts to fail) + */ + ret = mutex_lock_interruptible(&request_mutex); + if (ret < 0) + return ret; + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: priority remount in progress\n", + __func__); + spin_lock(&orangefs_superblocks_lock); + list_for_each_entry(orangefs_sb, &orangefs_superblocks, list) { + /* + * We have to drop the spinlock, so entries can be + * removed. They can't be freed, though, so we just + * keep the forward pointers and zero the back ones - + * that way we can get to the rest of the list. + */ + if (!orangefs_sb->list.prev) + continue; + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: Remounting SB %p\n", + __func__, + orangefs_sb); + + spin_unlock(&orangefs_superblocks_lock); + ret = orangefs_remount(orangefs_sb); + spin_lock(&orangefs_superblocks_lock); + if (ret) { + gossip_debug(GOSSIP_DEV_DEBUG, + "SB %p remount failed\n", + orangefs_sb); + break; + } + } + spin_unlock(&orangefs_superblocks_lock); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: priority remount complete\n", + __func__); + mutex_unlock(&request_mutex); + return ret; + + case ORANGEFS_DEV_UPSTREAM: + ret = copy_to_user((void __user *)arg, + &upstream_kmod, + sizeof(upstream_kmod)); + + if (ret != 0) + return -EIO; + else + return ret; + + case ORANGEFS_DEV_CLIENT_MASK: + ret = copy_from_user(&mask2_info, + (void __user *)arg, + sizeof(struct dev_mask2_info_s)); + + if (ret != 0) + return -EIO; + + client_debug_mask.mask1 = mask2_info.mask1_value; + client_debug_mask.mask2 = mask2_info.mask2_value; + + pr_info("%s: client debug mask has been been received " + ":%llx: :%llx:\n", + __func__, + (unsigned long long)client_debug_mask.mask1, + (unsigned long long)client_debug_mask.mask2); + + return ret; + + case ORANGEFS_DEV_CLIENT_STRING: + ret = copy_from_user(&client_debug_array_string, + (void __user *)arg, + ORANGEFS_MAX_DEBUG_STRING_LEN); + /* + * The real client-core makes an effort to ensure + * that actual strings that aren't too long to fit in + * this buffer is what we get here. We're going to use + * string functions on the stuff we got, so we'll make + * this extra effort to try and keep from + * flowing out of this buffer when we use the string + * functions, even if somehow the stuff we end up + * with here is garbage. + */ + client_debug_array_string[ORANGEFS_MAX_DEBUG_STRING_LEN - 1] = + '\0'; + + if (ret != 0) { + pr_info("%s: CLIENT_STRING: copy_from_user failed\n", + __func__); + return -EIO; + } + + pr_info("%s: client debug array string has been received.\n", + __func__); + + if (!help_string_initialized) { + + /* Free the "we don't know yet" default string... */ + kfree(debug_help_string); + + /* build a proper debug help string */ + if (orangefs_prepare_debugfs_help_string(0)) { + gossip_err("%s: no debug help string \n", + __func__); + return -EIO; + } + + /* Replace the boilerplate boot-time debug-help file. */ + debugfs_remove(help_file_dentry); + + help_file_dentry = + debugfs_create_file( + ORANGEFS_KMOD_DEBUG_HELP_FILE, + 0444, + debug_dir, + debug_help_string, + &debug_help_fops); + + if (!help_file_dentry) { + gossip_err("%s: debugfs_create_file failed for" + " :%s:!\n", + __func__, + ORANGEFS_KMOD_DEBUG_HELP_FILE); + return -EIO; + } + } + + debug_mask_to_string(&client_debug_mask, 1); + + debugfs_remove(client_debug_dentry); + + orangefs_client_debug_init(); + + help_string_initialized++; + + return ret; + + case ORANGEFS_DEV_DEBUG: + ret = copy_from_user(&mask_info, + (void __user *)arg, + sizeof(mask_info)); + + if (ret != 0) + return -EIO; + + if (mask_info.mask_type == KERNEL_MASK) { + if ((mask_info.mask_value == 0) + && (kernel_mask_set_mod_init)) { + /* + * the kernel debug mask was set when the + * kernel module was loaded; don't override + * it if the client-core was started without + * a value for ORANGEFS_KMODMASK. + */ + return 0; + } + debug_mask_to_string(&mask_info.mask_value, + mask_info.mask_type); + gossip_debug_mask = mask_info.mask_value; + pr_info("%s: kernel debug mask has been modified to " + ":%s: :%llx:\n", + __func__, + kernel_debug_string, + (unsigned long long)gossip_debug_mask); + } else if (mask_info.mask_type == CLIENT_MASK) { + debug_mask_to_string(&mask_info.mask_value, + mask_info.mask_type); + pr_info("%s: client debug mask has been modified to" + ":%s: :%llx:\n", + __func__, + client_debug_string, + llu(mask_info.mask_value)); + } else { + gossip_lerr("Invalid mask type....\n"); + return -EINVAL; + } + + return ret; + + default: + return -ENOIOCTLCMD; + } + return -ENOIOCTLCMD; +} + +static long orangefs_devreq_ioctl(struct file *file, + unsigned int command, unsigned long arg) +{ + long ret; + + /* Check for properly constructed commands */ + ret = check_ioctl_command(command); + if (ret < 0) + return (int)ret; + + return (int)dispatch_ioctl_command(command, arg); +} + +#ifdef CONFIG_COMPAT /* CONFIG_COMPAT is in .config */ + +/* Compat structure for the ORANGEFS_DEV_MAP ioctl */ +struct ORANGEFS_dev_map_desc32 { + compat_uptr_t ptr; + __s32 total_size; + __s32 size; + __s32 count; +}; + +static unsigned long translate_dev_map26(unsigned long args, long *error) +{ + struct ORANGEFS_dev_map_desc32 __user *p32 = (void __user *)args; + /* + * Depending on the architecture, allocate some space on the + * user-call-stack based on our expected layout. + */ + struct ORANGEFS_dev_map_desc __user *p = + compat_alloc_user_space(sizeof(*p)); + compat_uptr_t addr; + + *error = 0; + /* get the ptr from the 32 bit user-space */ + if (get_user(addr, &p32->ptr)) + goto err; + /* try to put that into a 64-bit layout */ + if (put_user(compat_ptr(addr), &p->ptr)) + goto err; + /* copy the remaining fields */ + if (copy_in_user(&p->total_size, &p32->total_size, sizeof(__s32))) + goto err; + if (copy_in_user(&p->size, &p32->size, sizeof(__s32))) + goto err; + if (copy_in_user(&p->count, &p32->count, sizeof(__s32))) + goto err; + return (unsigned long)p; +err: + *error = -EFAULT; + return 0; +} + +/* + * 32 bit user-space apps' ioctl handlers when kernel modules + * is compiled as a 64 bit one + */ +static long orangefs_devreq_compat_ioctl(struct file *filp, unsigned int cmd, + unsigned long args) +{ + long ret; + unsigned long arg = args; + + /* Check for properly constructed commands */ + ret = check_ioctl_command(cmd); + if (ret < 0) + return ret; + if (cmd == ORANGEFS_DEV_MAP) { + /* + * convert the arguments to what we expect internally + * in kernel space + */ + arg = translate_dev_map26(args, &ret); + if (ret < 0) { + gossip_err("Could not translate dev map\n"); + return ret; + } + } + /* no other ioctl requires translation */ + return dispatch_ioctl_command(cmd, arg); +} + +#endif /* CONFIG_COMPAT is in .config */ + +/* the assigned character device major number */ +static int orangefs_dev_major; + +/* + * Initialize orangefs device specific state: + * Must be called at module load time only + */ +int orangefs_dev_init(void) +{ + /* register orangefs-req device */ + orangefs_dev_major = register_chrdev(0, + ORANGEFS_REQDEVICE_NAME, + &orangefs_devreq_file_operations); + if (orangefs_dev_major < 0) { + gossip_debug(GOSSIP_DEV_DEBUG, + "Failed to register /dev/%s (error %d)\n", + ORANGEFS_REQDEVICE_NAME, orangefs_dev_major); + return orangefs_dev_major; + } + + gossip_debug(GOSSIP_DEV_DEBUG, + "*** /dev/%s character device registered ***\n", + ORANGEFS_REQDEVICE_NAME); + gossip_debug(GOSSIP_DEV_DEBUG, "'mknod /dev/%s c %d 0'.\n", + ORANGEFS_REQDEVICE_NAME, orangefs_dev_major); + return 0; +} + +void orangefs_dev_cleanup(void) +{ + unregister_chrdev(orangefs_dev_major, ORANGEFS_REQDEVICE_NAME); + gossip_debug(GOSSIP_DEV_DEBUG, + "*** /dev/%s character device unregistered ***\n", + ORANGEFS_REQDEVICE_NAME); +} + +static unsigned int orangefs_devreq_poll(struct file *file, + struct poll_table_struct *poll_table) +{ + int poll_revent_mask = 0; + + poll_wait(file, &orangefs_request_list_waitq, poll_table); + + if (!list_empty(&orangefs_request_list)) + poll_revent_mask |= POLL_IN; + return poll_revent_mask; +} + +const struct file_operations orangefs_devreq_file_operations = { + .owner = THIS_MODULE, + .read = orangefs_devreq_read, + .write_iter = orangefs_devreq_write_iter, + .open = orangefs_devreq_open, + .release = orangefs_devreq_release, + .unlocked_ioctl = orangefs_devreq_ioctl, + +#ifdef CONFIG_COMPAT /* CONFIG_COMPAT is in .config */ + .compat_ioctl = orangefs_devreq_compat_ioctl, +#endif + .poll = orangefs_devreq_poll +}; diff --git a/fs/orangefs/dir.c b/fs/orangefs/dir.c new file mode 100644 index 000000000000..ba7dec40771e --- /dev/null +++ b/fs/orangefs/dir.c @@ -0,0 +1,398 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +/* + * decode routine used by kmod to deal with the blob sent from + * userspace for readdirs. The blob contains zero or more of these + * sub-blobs: + * __u32 - represents length of the character string that follows. + * string - between 1 and ORANGEFS_NAME_MAX bytes long. + * padding - (if needed) to cause the __u32 plus the string to be + * eight byte aligned. + * khandle - sizeof(khandle) bytes. + */ +static long decode_dirents(char *ptr, size_t size, + struct orangefs_readdir_response_s *readdir) +{ + int i; + struct orangefs_readdir_response_s *rd = + (struct orangefs_readdir_response_s *) ptr; + char *buf = ptr; + int khandle_size = sizeof(struct orangefs_khandle); + size_t offset = offsetof(struct orangefs_readdir_response_s, + dirent_array); + /* 8 reflects eight byte alignment */ + int smallest_blob = khandle_size + 8; + __u32 len; + int aligned_len; + int sizeof_u32 = sizeof(__u32); + long ret; + + gossip_debug(GOSSIP_DIR_DEBUG, "%s: size:%zu:\n", __func__, size); + + /* size is = offset on empty dirs, > offset on non-empty dirs... */ + if (size < offset) { + gossip_err("%s: size:%zu: offset:%zu:\n", + __func__, + size, + offset); + ret = -EINVAL; + goto out; + } + + if ((size == offset) && (readdir->orangefs_dirent_outcount != 0)) { + gossip_err("%s: size:%zu: dirent_outcount:%d:\n", + __func__, + size, + readdir->orangefs_dirent_outcount); + ret = -EINVAL; + goto out; + } + + readdir->token = rd->token; + readdir->orangefs_dirent_outcount = rd->orangefs_dirent_outcount; + readdir->dirent_array = kcalloc(readdir->orangefs_dirent_outcount, + sizeof(*readdir->dirent_array), + GFP_KERNEL); + if (readdir->dirent_array == NULL) { + gossip_err("%s: kcalloc failed.\n", __func__); + ret = -ENOMEM; + goto out; + } + + buf += offset; + size -= offset; + + for (i = 0; i < readdir->orangefs_dirent_outcount; i++) { + if (size < smallest_blob) { + gossip_err("%s: size:%zu: smallest_blob:%d:\n", + __func__, + size, + smallest_blob); + ret = -EINVAL; + goto free; + } + + len = *(__u32 *)buf; + if ((len < 1) || (len > ORANGEFS_NAME_MAX)) { + gossip_err("%s: len:%d:\n", __func__, len); + ret = -EINVAL; + goto free; + } + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: size:%zu: len:%d:\n", + __func__, + size, + len); + + readdir->dirent_array[i].d_name = buf + sizeof_u32; + readdir->dirent_array[i].d_length = len; + + /* + * Calculate "aligned" length of this string and its + * associated __u32 descriptor. + */ + aligned_len = ((sizeof_u32 + len + 1) + 7) & ~7; + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: aligned_len:%d:\n", + __func__, + aligned_len); + + /* + * The end of the blob should coincide with the end + * of the last sub-blob. + */ + if (size < aligned_len + khandle_size) { + gossip_err("%s: ran off the end of the blob.\n", + __func__); + ret = -EINVAL; + goto free; + } + size -= aligned_len + khandle_size; + + buf += aligned_len; + + readdir->dirent_array[i].khandle = + *(struct orangefs_khandle *) buf; + buf += khandle_size; + } + ret = buf - ptr; + gossip_debug(GOSSIP_DIR_DEBUG, "%s: returning:%ld:\n", __func__, ret); + goto out; + +free: + kfree(readdir->dirent_array); + readdir->dirent_array = NULL; + +out: + return ret; +} + +/* + * Read directory entries from an instance of an open directory. + */ +static int orangefs_readdir(struct file *file, struct dir_context *ctx) +{ + int ret = 0; + int buffer_index; + /* + * ptoken supports Orangefs' distributed directory logic, added + * in 2.9.2. + */ + __u64 *ptoken = file->private_data; + __u64 pos = 0; + ino_t ino = 0; + struct dentry *dentry = file->f_path.dentry; + struct orangefs_kernel_op_s *new_op = NULL; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(dentry->d_inode); + int buffer_full = 0; + struct orangefs_readdir_response_s readdir_response; + void *dents_buf; + int i = 0; + int len = 0; + ino_t current_ino = 0; + char *current_entry = NULL; + long bytes_decoded; + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: ctx->pos:%lld, ptoken = %llu\n", + __func__, + lld(ctx->pos), + llu(*ptoken)); + + pos = (__u64) ctx->pos; + + /* are we done? */ + if (pos == ORANGEFS_READDIR_END) { + gossip_debug(GOSSIP_DIR_DEBUG, + "Skipping to termination path\n"); + return 0; + } + + gossip_debug(GOSSIP_DIR_DEBUG, + "orangefs_readdir called on %s (pos=%llu)\n", + dentry->d_name.name, llu(pos)); + + memset(&readdir_response, 0, sizeof(readdir_response)); + + new_op = op_alloc(ORANGEFS_VFS_OP_READDIR); + if (!new_op) + return -ENOMEM; + + /* + * Only the indices are shared. No memory is actually shared, but the + * mechanism is used. + */ + new_op->uses_shared_memory = 1; + new_op->upcall.req.readdir.refn = orangefs_inode->refn; + new_op->upcall.req.readdir.max_dirent_count = + ORANGEFS_MAX_DIRENT_COUNT_READDIR; + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: upcall.req.readdir.refn.khandle: %pU\n", + __func__, + &new_op->upcall.req.readdir.refn.khandle); + + new_op->upcall.req.readdir.token = *ptoken; + +get_new_buffer_index: + buffer_index = orangefs_readdir_index_get(); + if (buffer_index < 0) { + ret = buffer_index; + gossip_lerr("orangefs_readdir: orangefs_readdir_index_get() failure (%d)\n", + ret); + goto out_free_op; + } + new_op->upcall.req.readdir.buf_index = buffer_index; + + ret = service_operation(new_op, + "orangefs_readdir", + get_interruptible_flag(dentry->d_inode)); + + gossip_debug(GOSSIP_DIR_DEBUG, + "Readdir downcall status is %d. ret:%d\n", + new_op->downcall.status, + ret); + + orangefs_readdir_index_put(buffer_index); + + if (ret == -EAGAIN && op_state_purged(new_op)) { + /* Client-core indices are invalid after it restarted. */ + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: Getting new buffer_index for retry of readdir..\n", + __func__); + goto get_new_buffer_index; + } + + if (ret == -EIO && op_state_purged(new_op)) { + gossip_err("%s: Client is down. Aborting readdir call.\n", + __func__); + goto out_free_op; + } + + if (ret < 0 || new_op->downcall.status != 0) { + gossip_debug(GOSSIP_DIR_DEBUG, + "Readdir request failed. Status:%d\n", + new_op->downcall.status); + if (ret >= 0) + ret = new_op->downcall.status; + goto out_free_op; + } + + dents_buf = new_op->downcall.trailer_buf; + if (dents_buf == NULL) { + gossip_err("Invalid NULL buffer in readdir response\n"); + ret = -ENOMEM; + goto out_free_op; + } + + bytes_decoded = decode_dirents(dents_buf, new_op->downcall.trailer_size, + &readdir_response); + if (bytes_decoded < 0) { + ret = bytes_decoded; + gossip_err("Could not decode readdir from buffer %d\n", ret); + goto out_vfree; + } + + if (bytes_decoded != new_op->downcall.trailer_size) { + gossip_err("orangefs_readdir: # bytes decoded (%ld) " + "!= trailer size (%ld)\n", + bytes_decoded, + (long)new_op->downcall.trailer_size); + ret = -EINVAL; + goto out_destroy_handle; + } + + /* + * orangefs doesn't actually store dot and dot-dot, but + * we need to have them represented. + */ + if (pos == 0) { + ino = get_ino_from_khandle(dentry->d_inode); + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: calling dir_emit of \".\" with pos = %llu\n", + __func__, + llu(pos)); + ret = dir_emit(ctx, ".", 1, ino, DT_DIR); + pos += 1; + } + + if (pos == 1) { + ino = get_parent_ino_from_dentry(dentry); + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: calling dir_emit of \"..\" with pos = %llu\n", + __func__, + llu(pos)); + ret = dir_emit(ctx, "..", 2, ino, DT_DIR); + pos += 1; + } + + /* + * we stored ORANGEFS_ITERATE_NEXT in ctx->pos last time around + * to prevent "finding" dot and dot-dot on any iteration + * other than the first. + */ + if (ctx->pos == ORANGEFS_ITERATE_NEXT) + ctx->pos = 0; + + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: dirent_outcount:%d:\n", + __func__, + readdir_response.orangefs_dirent_outcount); + for (i = ctx->pos; + i < readdir_response.orangefs_dirent_outcount; + i++) { + len = readdir_response.dirent_array[i].d_length; + current_entry = readdir_response.dirent_array[i].d_name; + current_ino = orangefs_khandle_to_ino( + &readdir_response.dirent_array[i].khandle); + + gossip_debug(GOSSIP_DIR_DEBUG, + "calling dir_emit for %s with len %d" + ", ctx->pos %ld\n", + current_entry, + len, + (unsigned long)ctx->pos); + /* + * type is unknown. We don't return object type + * in the dirent_array. This leaves getdents + * clueless about type. + */ + ret = + dir_emit(ctx, current_entry, len, current_ino, DT_UNKNOWN); + if (!ret) + break; + ctx->pos++; + gossip_debug(GOSSIP_DIR_DEBUG, + "%s: ctx->pos:%lld\n", + __func__, + lld(ctx->pos)); + + } + + /* + * we ran all the way through the last batch, set up for + * getting another batch... + */ + if (ret) { + *ptoken = readdir_response.token; + ctx->pos = ORANGEFS_ITERATE_NEXT; + } + + /* + * Did we hit the end of the directory? + */ + if (readdir_response.token == ORANGEFS_READDIR_END && + !buffer_full) { + gossip_debug(GOSSIP_DIR_DEBUG, + "End of dir detected; setting ctx->pos to ORANGEFS_READDIR_END.\n"); + ctx->pos = ORANGEFS_READDIR_END; + } + +out_destroy_handle: + /* kfree(NULL) is safe */ + kfree(readdir_response.dirent_array); +out_vfree: + gossip_debug(GOSSIP_DIR_DEBUG, "vfree %p\n", dents_buf); + vfree(dents_buf); +out_free_op: + op_release(new_op); + gossip_debug(GOSSIP_DIR_DEBUG, "orangefs_readdir returning %d\n", ret); + return ret; +} + +static int orangefs_dir_open(struct inode *inode, struct file *file) +{ + __u64 *ptoken; + + file->private_data = kmalloc(sizeof(__u64), GFP_KERNEL); + if (!file->private_data) + return -ENOMEM; + + ptoken = file->private_data; + *ptoken = ORANGEFS_READDIR_START; + return 0; +} + +static int orangefs_dir_release(struct inode *inode, struct file *file) +{ + orangefs_flush_inode(inode); + kfree(file->private_data); + return 0; +} + +/** ORANGEFS implementation of VFS directory operations */ +const struct file_operations orangefs_dir_operations = { + .read = generic_read_dir, + .iterate = orangefs_readdir, + .open = orangefs_dir_open, + .release = orangefs_dir_release, +}; diff --git a/fs/orangefs/downcall.h b/fs/orangefs/downcall.h new file mode 100644 index 000000000000..66b99210f1f9 --- /dev/null +++ b/fs/orangefs/downcall.h @@ -0,0 +1,133 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Definitions of downcalls used in Linux kernel module. + */ + +#ifndef __DOWNCALL_H +#define __DOWNCALL_H + +/* + * Sanitized the device-client core interaction + * for clean 32-64 bit usage + */ +struct orangefs_io_response { + __s64 amt_complete; +}; + +struct orangefs_lookup_response { + struct orangefs_object_kref refn; +}; + +struct orangefs_create_response { + struct orangefs_object_kref refn; +}; + +struct orangefs_symlink_response { + struct orangefs_object_kref refn; +}; + +struct orangefs_getattr_response { + struct ORANGEFS_sys_attr_s attributes; + char link_target[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_mkdir_response { + struct orangefs_object_kref refn; +}; + +/* + * duplication of some system interface structures so that I don't have + * to allocate extra memory + */ +struct orangefs_dirent { + char *d_name; + int d_length; + struct orangefs_khandle khandle; +}; + +struct orangefs_statfs_response { + __s64 block_size; + __s64 blocks_total; + __s64 blocks_avail; + __s64 files_total; + __s64 files_avail; +}; + +struct orangefs_fs_mount_response { + __s32 fs_id; + __s32 id; + struct orangefs_khandle root_khandle; +}; + +/* the getxattr response is the attribute value */ +struct orangefs_getxattr_response { + __s32 val_sz; + __s32 __pad1; + char val[ORANGEFS_MAX_XATTR_VALUELEN]; +}; + +/* the listxattr response is an array of attribute names */ +struct orangefs_listxattr_response { + __s32 returned_count; + __s32 __pad1; + __u64 token; + char key[ORANGEFS_MAX_XATTR_LISTLEN * ORANGEFS_MAX_XATTR_NAMELEN]; + __s32 keylen; + __s32 __pad2; + __s32 lengths[ORANGEFS_MAX_XATTR_LISTLEN]; +}; + +struct orangefs_param_response { + __s64 value; +}; + +#define PERF_COUNT_BUF_SIZE 4096 +struct orangefs_perf_count_response { + char buffer[PERF_COUNT_BUF_SIZE]; +}; + +#define FS_KEY_BUF_SIZE 4096 +struct orangefs_fs_key_response { + __s32 fs_keylen; + __s32 __pad1; + char fs_key[FS_KEY_BUF_SIZE]; +}; + +struct orangefs_downcall_s { + __s32 type; + __s32 status; + /* currently trailer is used only by readdir */ + __s64 trailer_size; + char *trailer_buf; + + union { + struct orangefs_io_response io; + struct orangefs_lookup_response lookup; + struct orangefs_create_response create; + struct orangefs_symlink_response sym; + struct orangefs_getattr_response getattr; + struct orangefs_mkdir_response mkdir; + struct orangefs_statfs_response statfs; + struct orangefs_fs_mount_response fs_mount; + struct orangefs_getxattr_response getxattr; + struct orangefs_listxattr_response listxattr; + struct orangefs_param_response param; + struct orangefs_perf_count_response perf_count; + struct orangefs_fs_key_response fs_key; + } resp; +}; + +struct orangefs_readdir_response_s { + __u64 token; + __u64 directory_version; + __u32 __pad2; + __u32 orangefs_dirent_outcount; + struct orangefs_dirent *dirent_array; +}; + +#endif /* __DOWNCALL_H */ diff --git a/fs/orangefs/file.c b/fs/orangefs/file.c new file mode 100644 index 000000000000..ae92795ed965 --- /dev/null +++ b/fs/orangefs/file.c @@ -0,0 +1,717 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS file operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" +#include <linux/fs.h> +#include <linux/pagemap.h> + +/* + * Copy to client-core's address space from the buffers specified + * by the iovec upto total_size bytes. + * NOTE: the iovector can either contain addresses which + * can futher be kernel-space or user-space addresses. + * or it can pointers to struct page's + */ +static int precopy_buffers(int buffer_index, + struct iov_iter *iter, + size_t total_size) +{ + int ret = 0; + /* + * copy data from application/kernel by pulling it out + * of the iovec. + */ + + + if (total_size) { + ret = orangefs_bufmap_copy_from_iovec(iter, + buffer_index, + total_size); + if (ret < 0) + gossip_err("%s: Failed to copy-in buffers. Please make sure that the pvfs2-client is running. %ld\n", + __func__, + (long)ret); + } + + if (ret < 0) + gossip_err("%s: Failed to copy-in buffers. Please make sure that the pvfs2-client is running. %ld\n", + __func__, + (long)ret); + return ret; +} + +/* + * Copy from client-core's address space to the buffers specified + * by the iovec upto total_size bytes. + * NOTE: the iovector can either contain addresses which + * can futher be kernel-space or user-space addresses. + * or it can pointers to struct page's + */ +static int postcopy_buffers(int buffer_index, + struct iov_iter *iter, + size_t total_size) +{ + int ret = 0; + /* + * copy data to application/kernel by pushing it out to + * the iovec. NOTE; target buffers can be addresses or + * struct page pointers. + */ + if (total_size) { + ret = orangefs_bufmap_copy_to_iovec(iter, + buffer_index, + total_size); + if (ret < 0) + gossip_err("%s: Failed to copy-out buffers. Please make sure that the pvfs2-client is running (%ld)\n", + __func__, + (long)ret); + } + return ret; +} + +/* + * Post and wait for the I/O upcall to finish + */ +static ssize_t wait_for_direct_io(enum ORANGEFS_io_type type, struct inode *inode, + loff_t *offset, struct iov_iter *iter, + size_t total_size, loff_t readahead_size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_khandle *handle = &orangefs_inode->refn.khandle; + struct orangefs_kernel_op_s *new_op = NULL; + struct iov_iter saved = *iter; + int buffer_index = -1; + ssize_t ret; + + new_op = op_alloc(ORANGEFS_VFS_OP_FILE_IO); + if (!new_op) + return -ENOMEM; + + /* synchronous I/O */ + new_op->upcall.req.io.readahead_size = readahead_size; + new_op->upcall.req.io.io_type = type; + new_op->upcall.req.io.refn = orangefs_inode->refn; + +populate_shared_memory: + /* get a shared buffer index */ + buffer_index = orangefs_bufmap_get(); + if (buffer_index < 0) { + ret = buffer_index; + gossip_debug(GOSSIP_FILE_DEBUG, + "%s: orangefs_bufmap_get failure (%zd)\n", + __func__, ret); + goto out; + } + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): GET op %p -> buffer_index %d\n", + __func__, + handle, + new_op, + buffer_index); + + new_op->uses_shared_memory = 1; + new_op->upcall.req.io.buf_index = buffer_index; + new_op->upcall.req.io.count = total_size; + new_op->upcall.req.io.offset = *offset; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): offset: %llu total_size: %zd\n", + __func__, + handle, + llu(*offset), + total_size); + /* + * Stage 1: copy the buffers into client-core's address space + * precopy_buffers only pertains to writes. + */ + if (type == ORANGEFS_IO_WRITE) { + ret = precopy_buffers(buffer_index, + iter, + total_size); + if (ret < 0) + goto out; + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Calling post_io_request with tag (%llu)\n", + __func__, + handle, + llu(new_op->tag)); + + /* Stage 2: Service the I/O operation */ + ret = service_operation(new_op, + type == ORANGEFS_IO_WRITE ? + "file_write" : + "file_read", + get_interruptible_flag(inode)); + + /* + * If service_operation() returns -EAGAIN #and# the operation was + * purged from orangefs_request_list or htable_ops_in_progress, then + * we know that the client was restarted, causing the shared memory + * area to be wiped clean. To restart a write operation in this + * case, we must re-copy the data from the user's iovec to a NEW + * shared memory location. To restart a read operation, we must get + * a new shared memory location. + */ + if (ret == -EAGAIN && op_state_purged(new_op)) { + orangefs_bufmap_put(buffer_index); + buffer_index = -1; + if (type == ORANGEFS_IO_WRITE) + *iter = saved; + gossip_debug(GOSSIP_FILE_DEBUG, + "%s:going to repopulate_shared_memory.\n", + __func__); + goto populate_shared_memory; + } + + if (ret < 0) { + if (ret == -EINTR) { + /* + * We can't return EINTR if any data was written, + * it's not POSIX. It is minimally acceptable + * to give a partial write, the way NFS does. + * + * It would be optimal to return all or nothing, + * but if a userspace write is bigger than + * an IO buffer, and the interrupt occurs + * between buffer writes, that would not be + * possible. + */ + switch (new_op->op_state - OP_VFS_STATE_GIVEN_UP) { + /* + * If the op was waiting when the interrupt + * occurred, then the client-core did not + * trigger the write. + */ + case OP_VFS_STATE_WAITING: + if (*offset == 0) + ret = -EINTR; + else + ret = 0; + break; + /* + * If the op was in progress when the interrupt + * occurred, then the client-core was able to + * trigger the write. + */ + case OP_VFS_STATE_INPROGR: + ret = total_size; + break; + default: + gossip_err("%s: unexpected op state :%d:.\n", + __func__, + new_op->op_state); + ret = 0; + break; + } + gossip_debug(GOSSIP_FILE_DEBUG, + "%s: got EINTR, state:%d: %p\n", + __func__, + new_op->op_state, + new_op); + } else { + gossip_err("%s: error in %s handle %pU, returning %zd\n", + __func__, + type == ORANGEFS_IO_READ ? + "read from" : "write to", + handle, ret); + } + if (orangefs_cancel_op_in_progress(new_op)) + return ret; + + goto out; + } + + /* + * Stage 3: Post copy buffers from client-core's address space + * postcopy_buffers only pertains to reads. + */ + if (type == ORANGEFS_IO_READ) { + ret = postcopy_buffers(buffer_index, + iter, + new_op->downcall.resp.io.amt_complete); + if (ret < 0) + goto out; + } + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Amount %s, returned by the sys-io call:%d\n", + __func__, + handle, + type == ORANGEFS_IO_READ ? "read" : "written", + (int)new_op->downcall.resp.io.amt_complete); + + ret = new_op->downcall.resp.io.amt_complete; + +out: + if (buffer_index >= 0) { + orangefs_bufmap_put(buffer_index); + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): PUT buffer_index %d\n", + __func__, handle, buffer_index); + buffer_index = -1; + } + op_release(new_op); + return ret; +} + +/* + * Common entry point for read/write/readv/writev + * This function will dispatch it to either the direct I/O + * or buffered I/O path depending on the mount options and/or + * augmented/extended metadata attached to the file. + * Note: File extended attributes override any mount options. + */ +static ssize_t do_readv_writev(enum ORANGEFS_io_type type, struct file *file, + loff_t *offset, struct iov_iter *iter) +{ + struct inode *inode = file->f_mapping->host; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_khandle *handle = &orangefs_inode->refn.khandle; + size_t count = iov_iter_count(iter); + ssize_t total_count = 0; + ssize_t ret = -EINVAL; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s-BEGIN(%pU): count(%d) after estimate_max_iovecs.\n", + __func__, + handle, + (int)count); + + if (type == ORANGEFS_IO_WRITE) { + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): proceeding with offset : %llu, " + "size %d\n", + __func__, + handle, + llu(*offset), + (int)count); + } + + if (count == 0) { + ret = 0; + goto out; + } + + while (iov_iter_count(iter)) { + size_t each_count = iov_iter_count(iter); + size_t amt_complete; + + /* how much to transfer in this loop iteration */ + if (each_count > orangefs_bufmap_size_query()) + each_count = orangefs_bufmap_size_query(); + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): size of each_count(%d)\n", + __func__, + handle, + (int)each_count); + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): BEFORE wait_for_io: offset is %d\n", + __func__, + handle, + (int)*offset); + + ret = wait_for_direct_io(type, inode, offset, iter, + each_count, 0); + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): return from wait_for_io:%d\n", + __func__, + handle, + (int)ret); + + if (ret < 0) + goto out; + + *offset += ret; + total_count += ret; + amt_complete = ret; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): AFTER wait_for_io: offset is %d\n", + __func__, + handle, + (int)*offset); + + /* + * if we got a short I/O operations, + * fall out and return what we got so far + */ + if (amt_complete < each_count) + break; + } /*end while */ + +out: + if (total_count > 0) + ret = total_count; + if (ret > 0) { + if (type == ORANGEFS_IO_READ) { + file_accessed(file); + } else { + SetMtimeFlag(orangefs_inode); + inode->i_mtime = CURRENT_TIME; + mark_inode_dirty_sync(inode); + } + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Value(%d) returned.\n", + __func__, + handle, + (int)ret); + + return ret; +} + +/* + * Read data from a specified offset in a file (referenced by inode). + * Data may be placed either in a user or kernel buffer. + */ +ssize_t orangefs_inode_read(struct inode *inode, + struct iov_iter *iter, + loff_t *offset, + loff_t readahead_size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + size_t count = iov_iter_count(iter); + size_t bufmap_size; + ssize_t ret = -EINVAL; + + g_orangefs_stats.reads++; + + bufmap_size = orangefs_bufmap_size_query(); + if (count > bufmap_size) { + gossip_debug(GOSSIP_FILE_DEBUG, + "%s: count is too large (%zd/%zd)!\n", + __func__, count, bufmap_size); + return -EINVAL; + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU) %zd@%llu\n", + __func__, + &orangefs_inode->refn.khandle, + count, + llu(*offset)); + + ret = wait_for_direct_io(ORANGEFS_IO_READ, inode, offset, iter, + count, readahead_size); + if (ret > 0) + *offset += ret; + + gossip_debug(GOSSIP_FILE_DEBUG, + "%s(%pU): Value(%zd) returned.\n", + __func__, + &orangefs_inode->refn.khandle, + ret); + + return ret; +} + +static ssize_t orangefs_file_read_iter(struct kiocb *iocb, struct iov_iter *iter) +{ + struct file *file = iocb->ki_filp; + loff_t pos = *(&iocb->ki_pos); + ssize_t rc = 0; + + BUG_ON(iocb->private); + + gossip_debug(GOSSIP_FILE_DEBUG, "orangefs_file_read_iter\n"); + + g_orangefs_stats.reads++; + + rc = do_readv_writev(ORANGEFS_IO_READ, file, &pos, iter); + iocb->ki_pos = pos; + + return rc; +} + +static ssize_t orangefs_file_write_iter(struct kiocb *iocb, struct iov_iter *iter) +{ + struct file *file = iocb->ki_filp; + loff_t pos; + ssize_t rc; + + BUG_ON(iocb->private); + + gossip_debug(GOSSIP_FILE_DEBUG, "orangefs_file_write_iter\n"); + + mutex_lock(&file->f_mapping->host->i_mutex); + + /* Make sure generic_write_checks sees an up to date inode size. */ + if (file->f_flags & O_APPEND) { + rc = orangefs_inode_getattr(file->f_mapping->host, 0, 1); + if (rc == -ESTALE) + rc = -EIO; + if (rc) { + gossip_err("%s: orangefs_inode_getattr failed, " + "rc:%zd:.\n", __func__, rc); + goto out; + } + } + + if (file->f_pos > i_size_read(file->f_mapping->host)) + orangefs_i_size_write(file->f_mapping->host, file->f_pos); + + rc = generic_write_checks(iocb, iter); + + if (rc <= 0) { + gossip_err("%s: generic_write_checks failed, rc:%zd:.\n", + __func__, rc); + goto out; + } + + /* + * if we are appending, generic_write_checks would have updated + * pos to the end of the file, so we will wait till now to set + * pos... + */ + pos = *(&iocb->ki_pos); + + rc = do_readv_writev(ORANGEFS_IO_WRITE, + file, + &pos, + iter); + if (rc < 0) { + gossip_err("%s: do_readv_writev failed, rc:%zd:.\n", + __func__, rc); + goto out; + } + + iocb->ki_pos = pos; + g_orangefs_stats.writes++; + +out: + + mutex_unlock(&file->f_mapping->host->i_mutex); + return rc; +} + +/* + * Perform a miscellaneous operation on a file. + */ +static long orangefs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) +{ + int ret = -ENOTTY; + __u64 val = 0; + unsigned long uval; + + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_ioctl: called with cmd %d\n", + cmd); + + /* + * we understand some general ioctls on files, such as the immutable + * and append flags + */ + if (cmd == FS_IOC_GETFLAGS) { + val = 0; + ret = orangefs_inode_getxattr(file_inode(file), + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + "user.pvfs2.meta_hint", + &val, sizeof(val)); + if (ret < 0 && ret != -ENODATA) + return ret; + else if (ret == -ENODATA) + val = 0; + uval = val; + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_ioctl: FS_IOC_GETFLAGS: %llu\n", + (unsigned long long)uval); + return put_user(uval, (int __user *)arg); + } else if (cmd == FS_IOC_SETFLAGS) { + ret = 0; + if (get_user(uval, (int __user *)arg)) + return -EFAULT; + /* + * ORANGEFS_MIRROR_FL is set internally when the mirroring mode + * is turned on for a file. The user is not allowed to turn + * on this bit, but the bit is present if the user first gets + * the flags and then updates the flags with some new + * settings. So, we ignore it in the following edit. bligon. + */ + if ((uval & ~ORANGEFS_MIRROR_FL) & + (~(FS_IMMUTABLE_FL | FS_APPEND_FL | FS_NOATIME_FL))) { + gossip_err("orangefs_ioctl: the FS_IOC_SETFLAGS only supports setting one of FS_IMMUTABLE_FL|FS_APPEND_FL|FS_NOATIME_FL\n"); + return -EINVAL; + } + val = uval; + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_ioctl: FS_IOC_SETFLAGS: %llu\n", + (unsigned long long)val); + ret = orangefs_inode_setxattr(file_inode(file), + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + "user.pvfs2.meta_hint", + &val, sizeof(val), 0); + } + + return ret; +} + +/* + * Memory map a region of a file. + */ +static int orangefs_file_mmap(struct file *file, struct vm_area_struct *vma) +{ + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_file_mmap: called on %s\n", + (file ? + (char *)file->f_path.dentry->d_name.name : + (char *)"Unknown")); + + /* set the sequential readahead hint */ + vma->vm_flags |= VM_SEQ_READ; + vma->vm_flags &= ~VM_RAND_READ; + + /* Use readonly mmap since we cannot support writable maps. */ + return generic_file_readonly_mmap(file, vma); +} + +#define mapping_nrpages(idata) ((idata)->nrpages) + +/* + * Called to notify the module that there are no more references to + * this file (i.e. no processes have it open). + * + * \note Not called when each file is closed. + */ +static int orangefs_file_release(struct inode *inode, struct file *file) +{ + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_file_release: called on %s\n", + file->f_path.dentry->d_name.name); + + orangefs_flush_inode(inode); + + /* + * remove all associated inode pages from the page cache and mmap + * readahead cache (if any); this forces an expensive refresh of + * data for the next caller of mmap (or 'get_block' accesses) + */ + if (file->f_path.dentry->d_inode && + file->f_path.dentry->d_inode->i_mapping && + mapping_nrpages(&file->f_path.dentry->d_inode->i_data)) + truncate_inode_pages(file->f_path.dentry->d_inode->i_mapping, + 0); + return 0; +} + +/* + * Push all data for a specific file onto permanent storage. + */ +static int orangefs_fsync(struct file *file, + loff_t start, + loff_t end, + int datasync) +{ + int ret = -EINVAL; + struct orangefs_inode_s *orangefs_inode = + ORANGEFS_I(file->f_path.dentry->d_inode); + struct orangefs_kernel_op_s *new_op = NULL; + + /* required call */ + filemap_write_and_wait_range(file->f_mapping, start, end); + + new_op = op_alloc(ORANGEFS_VFS_OP_FSYNC); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.fsync.refn = orangefs_inode->refn; + + ret = service_operation(new_op, + "orangefs_fsync", + get_interruptible_flag(file->f_path.dentry->d_inode)); + + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_fsync got return value of %d\n", + ret); + + op_release(new_op); + + orangefs_flush_inode(file->f_path.dentry->d_inode); + return ret; +} + +/* + * Change the file pointer position for an instance of an open file. + * + * \note If .llseek is overriden, we must acquire lock as described in + * Documentation/filesystems/Locking. + * + * Future upgrade could support SEEK_DATA and SEEK_HOLE but would + * require much changes to the FS + */ +static loff_t orangefs_file_llseek(struct file *file, loff_t offset, int origin) +{ + int ret = -EINVAL; + struct inode *inode = file_inode(file); + + if (origin == SEEK_END) { + /* + * revalidate the inode's file size. + * NOTE: We are only interested in file size here, + * so we set mask accordingly. + */ + ret = orangefs_inode_getattr(file->f_mapping->host, 0, 1); + if (ret == -ESTALE) + ret = -EIO; + if (ret) { + gossip_debug(GOSSIP_FILE_DEBUG, + "%s:%s:%d calling make bad inode\n", + __FILE__, + __func__, + __LINE__); + return ret; + } + } + + gossip_debug(GOSSIP_FILE_DEBUG, + "orangefs_file_llseek: offset is %ld | origin is %d" + " | inode size is %lu\n", + (long)offset, + origin, + (unsigned long)i_size_read(inode)); + + return generic_file_llseek(file, offset, origin); +} + +/* + * Support local locks (locks that only this kernel knows about) + * if Orangefs was mounted -o local_lock. + */ +static int orangefs_lock(struct file *filp, int cmd, struct file_lock *fl) +{ + int rc = -EINVAL; + + if (ORANGEFS_SB(filp->f_inode->i_sb)->flags & ORANGEFS_OPT_LOCAL_LOCK) { + if (cmd == F_GETLK) { + rc = 0; + posix_test_lock(filp, fl); + } else { + rc = posix_lock_file(filp, fl, NULL); + } + } + + return rc; +} + +/** ORANGEFS implementation of VFS file operations */ +const struct file_operations orangefs_file_operations = { + .llseek = orangefs_file_llseek, + .read_iter = orangefs_file_read_iter, + .write_iter = orangefs_file_write_iter, + .lock = orangefs_lock, + .unlocked_ioctl = orangefs_ioctl, + .mmap = orangefs_file_mmap, + .open = generic_file_open, + .release = orangefs_file_release, + .fsync = orangefs_fsync, +}; diff --git a/fs/orangefs/inode.c b/fs/orangefs/inode.c new file mode 100644 index 000000000000..2382e267b49e --- /dev/null +++ b/fs/orangefs/inode.c @@ -0,0 +1,475 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS inode operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +static int read_one_page(struct page *page) +{ + int ret; + int max_block; + ssize_t bytes_read = 0; + struct inode *inode = page->mapping->host; + const __u32 blocksize = PAGE_CACHE_SIZE; /* inode->i_blksize */ + const __u32 blockbits = PAGE_CACHE_SHIFT; /* inode->i_blkbits */ + struct iov_iter to; + struct bio_vec bv = {.bv_page = page, .bv_len = PAGE_SIZE}; + + iov_iter_bvec(&to, ITER_BVEC | READ, &bv, 1, PAGE_SIZE); + + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_readpage called with page %p\n", + page); + + max_block = ((inode->i_size / blocksize) + 1); + + if (page->index < max_block) { + loff_t blockptr_offset = (((loff_t) page->index) << blockbits); + + bytes_read = orangefs_inode_read(inode, + &to, + &blockptr_offset, + inode->i_size); + } + /* this will only zero remaining unread portions of the page data */ + iov_iter_zero(~0U, &to); + /* takes care of potential aliasing */ + flush_dcache_page(page); + if (bytes_read < 0) { + ret = bytes_read; + SetPageError(page); + } else { + SetPageUptodate(page); + if (PageError(page)) + ClearPageError(page); + ret = 0; + } + /* unlock the page after the ->readpage() routine completes */ + unlock_page(page); + return ret; +} + +static int orangefs_readpage(struct file *file, struct page *page) +{ + return read_one_page(page); +} + +static int orangefs_readpages(struct file *file, + struct address_space *mapping, + struct list_head *pages, + unsigned nr_pages) +{ + int page_idx; + int ret; + + gossip_debug(GOSSIP_INODE_DEBUG, "orangefs_readpages called\n"); + + for (page_idx = 0; page_idx < nr_pages; page_idx++) { + struct page *page; + + page = list_entry(pages->prev, struct page, lru); + list_del(&page->lru); + if (!add_to_page_cache(page, + mapping, + page->index, + GFP_KERNEL)) { + ret = read_one_page(page); + gossip_debug(GOSSIP_INODE_DEBUG, + "failure adding page to cache, read_one_page returned: %d\n", + ret); + } else { + page_cache_release(page); + } + } + BUG_ON(!list_empty(pages)); + return 0; +} + +static void orangefs_invalidatepage(struct page *page, + unsigned int offset, + unsigned int length) +{ + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_invalidatepage called on page %p " + "(offset is %u)\n", + page, + offset); + + ClearPageUptodate(page); + ClearPageMappedToDisk(page); + return; + +} + +static int orangefs_releasepage(struct page *page, gfp_t foo) +{ + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_releasepage called on page %p\n", + page); + return 0; +} + +/* + * Having a direct_IO entry point in the address_space_operations + * struct causes the kernel to allows us to use O_DIRECT on + * open. Nothing will ever call this thing, but in the future we + * will need to be able to use O_DIRECT on open in order to support + * AIO. Modeled after NFS, they do this too. + */ +/* + * static ssize_t orangefs_direct_IO(int rw, + * struct kiocb *iocb, + * struct iov_iter *iter, + * loff_t offset) + *{ + * gossip_debug(GOSSIP_INODE_DEBUG, + * "orangefs_direct_IO: %s\n", + * iocb->ki_filp->f_path.dentry->d_name.name); + * + * return -EINVAL; + *} + */ + +struct backing_dev_info orangefs_backing_dev_info = { + .name = "orangefs", + .ra_pages = 0, + .capabilities = BDI_CAP_NO_ACCT_DIRTY | BDI_CAP_NO_WRITEBACK, +}; + +/** ORANGEFS2 implementation of address space operations */ +const struct address_space_operations orangefs_address_operations = { + .readpage = orangefs_readpage, + .readpages = orangefs_readpages, + .invalidatepage = orangefs_invalidatepage, + .releasepage = orangefs_releasepage, +/* .direct_IO = orangefs_direct_IO */ +}; + +static int orangefs_setattr_size(struct inode *inode, struct iattr *iattr) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + loff_t orig_size; + int ret = -EINVAL; + + gossip_debug(GOSSIP_INODE_DEBUG, + "%s: %pU: Handle is %pU | fs_id %d | size is %llu\n", + __func__, + get_khandle_from_ino(inode), + &orangefs_inode->refn.khandle, + orangefs_inode->refn.fs_id, + iattr->ia_size); + + /* Ensure that we have a up to date size, so we know if it changed. */ + ret = orangefs_inode_getattr(inode, 0, 1); + if (ret == -ESTALE) + ret = -EIO; + if (ret) { + gossip_err("%s: orangefs_inode_getattr failed, ret:%d:.\n", + __func__, ret); + return ret; + } + orig_size = i_size_read(inode); + + truncate_setsize(inode, iattr->ia_size); + + new_op = op_alloc(ORANGEFS_VFS_OP_TRUNCATE); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.truncate.refn = orangefs_inode->refn; + new_op->upcall.req.truncate.size = (__s64) iattr->ia_size; + + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + + /* + * the truncate has no downcall members to retrieve, but + * the status value tells us if it went through ok or not + */ + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs: orangefs_truncate got return value of %d\n", + ret); + + op_release(new_op); + + if (ret != 0) + return ret; + + /* + * Only change the c/mtime if we are changing the size or we are + * explicitly asked to change it. This handles the semantic difference + * between truncate() and ftruncate() as implemented in the VFS. + * + * The regular truncate() case without ATTR_CTIME and ATTR_MTIME is a + * special case where we need to update the times despite not having + * these flags set. For all other operations the VFS set these flags + * explicitly if it wants a timestamp update. + */ + if (orig_size != i_size_read(inode) && + !(iattr->ia_valid & (ATTR_CTIME | ATTR_MTIME))) { + iattr->ia_ctime = iattr->ia_mtime = + current_fs_time(inode->i_sb); + iattr->ia_valid |= ATTR_CTIME | ATTR_MTIME; + } + + return ret; +} + +/* + * Change attributes of an object referenced by dentry. + */ +int orangefs_setattr(struct dentry *dentry, struct iattr *iattr) +{ + int ret = -EINVAL; + struct inode *inode = dentry->d_inode; + + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_setattr: called on %s\n", + dentry->d_name.name); + + ret = inode_change_ok(inode, iattr); + if (ret) + goto out; + + if ((iattr->ia_valid & ATTR_SIZE) && + iattr->ia_size != i_size_read(inode)) { + ret = orangefs_setattr_size(inode, iattr); + if (ret) + goto out; + } + + setattr_copy(inode, iattr); + mark_inode_dirty(inode); + + ret = orangefs_inode_setattr(inode, iattr); + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_setattr: inode_setattr returned %d\n", + ret); + + if (!ret && (iattr->ia_valid & ATTR_MODE)) + /* change mod on a file that has ACLs */ + ret = posix_acl_chmod(inode, inode->i_mode); + +out: + gossip_debug(GOSSIP_INODE_DEBUG, "orangefs_setattr: returning %d\n", ret); + return ret; +} + +/* + * Obtain attributes of an object given a dentry + */ +int orangefs_getattr(struct vfsmount *mnt, + struct dentry *dentry, + struct kstat *kstat) +{ + int ret = -ENOENT; + struct inode *inode = dentry->d_inode; + struct orangefs_inode_s *orangefs_inode = NULL; + + gossip_debug(GOSSIP_INODE_DEBUG, + "orangefs_getattr: called on %s\n", + dentry->d_name.name); + + ret = orangefs_inode_getattr(inode, 0, 1); + if (ret == 0) { + generic_fillattr(inode, kstat); + + /* override block size reported to stat */ + orangefs_inode = ORANGEFS_I(inode); + kstat->blksize = orangefs_inode->blksize; + } + return ret; +} + +int orangefs_permission(struct inode *inode, int mask) +{ + int ret; + + if (mask & MAY_NOT_BLOCK) + return -ECHILD; + + gossip_debug(GOSSIP_INODE_DEBUG, "%s: refreshing\n", __func__); + + /* Make sure the permission (and other common attrs) are up to date. */ + ret = orangefs_inode_getattr(inode, 0, 0); + if (ret < 0) + return ret; + + return generic_permission(inode, mask); +} + +/* ORANGEDS2 implementation of VFS inode operations for files */ +struct inode_operations orangefs_file_inode_operations = { + .get_acl = orangefs_get_acl, + .set_acl = orangefs_set_acl, + .setattr = orangefs_setattr, + .getattr = orangefs_getattr, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .listxattr = orangefs_listxattr, + .removexattr = generic_removexattr, + .permission = orangefs_permission, +}; + +static int orangefs_init_iops(struct inode *inode) +{ + inode->i_mapping->a_ops = &orangefs_address_operations; + + switch (inode->i_mode & S_IFMT) { + case S_IFREG: + inode->i_op = &orangefs_file_inode_operations; + inode->i_fop = &orangefs_file_operations; + inode->i_blkbits = PAGE_CACHE_SHIFT; + break; + case S_IFLNK: + inode->i_op = &orangefs_symlink_inode_operations; + break; + case S_IFDIR: + inode->i_op = &orangefs_dir_inode_operations; + inode->i_fop = &orangefs_dir_operations; + break; + default: + gossip_debug(GOSSIP_INODE_DEBUG, + "%s: unsupported mode\n", + __func__); + return -EINVAL; + } + + return 0; +} + +/* + * Given a ORANGEFS object identifier (fsid, handle), convert it into a ino_t type + * that will be used as a hash-index from where the handle will + * be searched for in the VFS hash table of inodes. + */ +static inline ino_t orangefs_handle_hash(struct orangefs_object_kref *ref) +{ + if (!ref) + return 0; + return orangefs_khandle_to_ino(&(ref->khandle)); +} + +/* + * Called to set up an inode from iget5_locked. + */ +static int orangefs_set_inode(struct inode *inode, void *data) +{ + struct orangefs_object_kref *ref = (struct orangefs_object_kref *) data; + ORANGEFS_I(inode)->refn.fs_id = ref->fs_id; + ORANGEFS_I(inode)->refn.khandle = ref->khandle; + return 0; +} + +/* + * Called to determine if handles match. + */ +static int orangefs_test_inode(struct inode *inode, void *data) +{ + struct orangefs_object_kref *ref = (struct orangefs_object_kref *) data; + struct orangefs_inode_s *orangefs_inode = NULL; + + orangefs_inode = ORANGEFS_I(inode); + return (!ORANGEFS_khandle_cmp(&(orangefs_inode->refn.khandle), &(ref->khandle)) + && orangefs_inode->refn.fs_id == ref->fs_id); +} + +/* + * Front-end to lookup the inode-cache maintained by the VFS using the ORANGEFS + * file handle. + * + * @sb: the file system super block instance. + * @ref: The ORANGEFS object for which we are trying to locate an inode structure. + */ +struct inode *orangefs_iget(struct super_block *sb, struct orangefs_object_kref *ref) +{ + struct inode *inode = NULL; + unsigned long hash; + int error; + + hash = orangefs_handle_hash(ref); + inode = iget5_locked(sb, hash, orangefs_test_inode, orangefs_set_inode, ref); + if (!inode || !(inode->i_state & I_NEW)) + return inode; + + error = orangefs_inode_getattr(inode, 1, 0); + if (error) { + iget_failed(inode); + return ERR_PTR(error); + } + + inode->i_ino = hash; /* needed for stat etc */ + orangefs_init_iops(inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_INODE_DEBUG, + "iget handle %pU, fsid %d hash %ld i_ino %lu\n", + &ref->khandle, + ref->fs_id, + hash, + inode->i_ino); + + return inode; +} + +/* + * Allocate an inode for a newly created file and insert it into the inode hash. + */ +struct inode *orangefs_new_inode(struct super_block *sb, struct inode *dir, + int mode, dev_t dev, struct orangefs_object_kref *ref) +{ + unsigned long hash = orangefs_handle_hash(ref); + struct inode *inode; + int error; + + gossip_debug(GOSSIP_INODE_DEBUG, + "%s:(sb is %p | MAJOR(dev)=%u | MINOR(dev)=%u mode=%o)\n", + __func__, + sb, + MAJOR(dev), + MINOR(dev), + mode); + + inode = new_inode(sb); + if (!inode) + return NULL; + + orangefs_set_inode(inode, ref); + inode->i_ino = hash; /* needed for stat etc */ + + error = orangefs_inode_getattr(inode, 1, 0); + if (error) + goto out_iput; + + orangefs_init_iops(inode); + + inode->i_mode = mode; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); + inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; + inode->i_size = PAGE_CACHE_SIZE; + inode->i_rdev = dev; + + error = insert_inode_locked4(inode, hash, orangefs_test_inode, ref); + if (error < 0) + goto out_iput; + + gossip_debug(GOSSIP_INODE_DEBUG, + "Initializing ACL's for inode %pU\n", + get_khandle_from_ino(inode)); + orangefs_init_acl(inode, dir); + return inode; + +out_iput: + iput(inode); + return ERR_PTR(error); +} diff --git a/fs/orangefs/namei.c b/fs/orangefs/namei.c new file mode 100644 index 000000000000..5a60c508af4e --- /dev/null +++ b/fs/orangefs/namei.c @@ -0,0 +1,462 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS namei operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" + +/* + * Get a newly allocated inode to go with a negative dentry. + */ +static int orangefs_create(struct inode *dir, + struct dentry *dentry, + umode_t mode, + bool exclusive) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, "%s: %s\n", + __func__, + dentry->d_name.name); + + new_op = op_alloc(ORANGEFS_VFS_OP_CREATE); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.create.parent_refn = parent->refn; + + fill_default_sys_attrs(new_op->upcall.req.create.attributes, + ORANGEFS_TYPE_METAFILE, mode); + + strncpy(new_op->upcall.req.create.d_name, + dentry->d_name.name, ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: %s: handle:%pU: fsid:%d: new_op:%p: ret:%d:\n", + __func__, + dentry->d_name.name, + &new_op->downcall.resp.create.refn.khandle, + new_op->downcall.resp.create.refn.fs_id, + new_op, + ret); + + if (ret < 0) + goto out; + + inode = orangefs_new_inode(dir->i_sb, dir, S_IFREG | mode, 0, + &new_op->downcall.resp.create.refn); + if (IS_ERR(inode)) { + gossip_err("%s: Failed to allocate inode for file :%s:\n", + __func__, + dentry->d_name.name); + ret = PTR_ERR(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: Assigned inode :%pU: for file :%s:\n", + __func__, + get_khandle_from_ino(inode), + dentry->d_name.name); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: dentry instantiated for %s\n", + __func__, + dentry->d_name.name); + + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); + ret = 0; +out: + op_release(new_op); + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: %s: returning %d\n", + __func__, + dentry->d_name.name, + ret); + return ret; +} + +/* + * Attempt to resolve an object name (dentry->d_name), parent handle, and + * fsid into a handle for the object. + */ +static struct dentry *orangefs_lookup(struct inode *dir, struct dentry *dentry, + unsigned int flags) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + struct dentry *res; + int ret = -EINVAL; + + /* + * in theory we could skip a lookup here (if the intent is to + * create) in order to avoid a potentially failed lookup, but + * leaving it in can skip a valid lookup and try to create a file + * that already exists (e.g. the vfs already handles checking for + * -EEXIST on O_EXCL opens, which is broken if we skip this lookup + * in the create path) + */ + gossip_debug(GOSSIP_NAME_DEBUG, "%s called on %s\n", + __func__, dentry->d_name.name); + + if (dentry->d_name.len > (ORANGEFS_NAME_MAX - 1)) + return ERR_PTR(-ENAMETOOLONG); + + new_op = op_alloc(ORANGEFS_VFS_OP_LOOKUP); + if (!new_op) + return ERR_PTR(-ENOMEM); + + new_op->upcall.req.lookup.sym_follow = ORANGEFS_LOOKUP_LINK_NO_FOLLOW; + + gossip_debug(GOSSIP_NAME_DEBUG, "%s:%s:%d using parent %pU\n", + __FILE__, + __func__, + __LINE__, + &parent->refn.khandle); + new_op->upcall.req.lookup.parent_refn = parent->refn; + + strncpy(new_op->upcall.req.lookup.d_name, dentry->d_name.name, + ORANGEFS_NAME_MAX); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: doing lookup on %s under %pU,%d\n", + __func__, + new_op->upcall.req.lookup.d_name, + &new_op->upcall.req.lookup.parent_refn.khandle, + new_op->upcall.req.lookup.parent_refn.fs_id); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Lookup Got %pU, fsid %d (ret=%d)\n", + &new_op->downcall.resp.lookup.refn.khandle, + new_op->downcall.resp.lookup.refn.fs_id, + ret); + + if (ret < 0) { + if (ret == -ENOENT) { + /* + * if no inode was found, add a negative dentry to + * dcache anyway; if we don't, we don't hold expected + * lookup semantics and we most noticeably break + * during directory renames. + * + * however, if the operation failed or exited, do not + * add the dentry (e.g. in the case that a touch is + * issued on a file that already exists that was + * interrupted during this lookup -- no need to add + * another negative dentry for an existing file) + */ + + gossip_debug(GOSSIP_NAME_DEBUG, + "orangefs_lookup: Adding *negative* dentry " + "%p for %s\n", + dentry, + dentry->d_name.name); + + d_add(dentry, NULL); + res = NULL; + goto out; + } + + /* must be a non-recoverable error */ + res = ERR_PTR(ret); + goto out; + } + + inode = orangefs_iget(dir->i_sb, &new_op->downcall.resp.lookup.refn); + if (IS_ERR(inode)) { + gossip_debug(GOSSIP_NAME_DEBUG, + "error %ld from iget\n", PTR_ERR(inode)); + res = ERR_CAST(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s:%s:%d " + "Found good inode [%lu] with count [%d]\n", + __FILE__, + __func__, + __LINE__, + inode->i_ino, + (int)atomic_read(&inode->i_count)); + + /* update dentry/inode pair into dcache */ + res = d_splice_alias(inode, dentry); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Lookup success (inode ct = %d)\n", + (int)atomic_read(&inode->i_count)); +out: + op_release(new_op); + return res; +} + +/* return 0 on success; non-zero otherwise */ +static int orangefs_unlink(struct inode *dir, struct dentry *dentry) +{ + struct inode *inode = dentry->d_inode; + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: called on %s\n" + " (inode %pU): Parent is %pU | fs_id %d\n", + __func__, + dentry->d_name.name, + get_khandle_from_ino(inode), + &parent->refn.khandle, + parent->refn.fs_id); + + new_op = op_alloc(ORANGEFS_VFS_OP_REMOVE); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.remove.parent_refn = parent->refn; + strncpy(new_op->upcall.req.remove.d_name, dentry->d_name.name, + ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, "orangefs_unlink", + get_interruptible_flag(inode)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: service_operation returned:%d:\n", + __func__, + ret); + + op_release(new_op); + + if (!ret) { + drop_nlink(inode); + + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); + } + return ret; +} + +static int orangefs_symlink(struct inode *dir, + struct dentry *dentry, + const char *symname) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + int mode = 755; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, "%s: called\n", __func__); + + if (!symname) + return -EINVAL; + + if (strlen(symname)+1 > ORANGEFS_NAME_MAX) + return -ENAMETOOLONG; + + new_op = op_alloc(ORANGEFS_VFS_OP_SYMLINK); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.sym.parent_refn = parent->refn; + + fill_default_sys_attrs(new_op->upcall.req.sym.attributes, + ORANGEFS_TYPE_SYMLINK, + mode); + + strncpy(new_op->upcall.req.sym.entry_name, + dentry->d_name.name, + ORANGEFS_NAME_MAX); + strncpy(new_op->upcall.req.sym.target, symname, ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Symlink Got ORANGEFS handle %pU on fsid %d (ret=%d)\n", + &new_op->downcall.resp.sym.refn.khandle, + new_op->downcall.resp.sym.refn.fs_id, ret); + + if (ret < 0) { + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: failed with error code %d\n", + __func__, ret); + goto out; + } + + inode = orangefs_new_inode(dir->i_sb, dir, S_IFLNK | mode, 0, + &new_op->downcall.resp.sym.refn); + if (IS_ERR(inode)) { + gossip_err + ("*** Failed to allocate orangefs symlink inode\n"); + ret = PTR_ERR(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "Assigned symlink inode new number of %pU\n", + get_khandle_from_ino(inode)); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Inode (Symlink) %pU -> %s\n", + get_khandle_from_ino(inode), + dentry->d_name.name); + + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); + ret = 0; +out: + op_release(new_op); + return ret; +} + +static int orangefs_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + struct orangefs_inode_s *parent = ORANGEFS_I(dir); + struct orangefs_kernel_op_s *new_op; + struct inode *inode; + int ret; + + new_op = op_alloc(ORANGEFS_VFS_OP_MKDIR); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.mkdir.parent_refn = parent->refn; + + fill_default_sys_attrs(new_op->upcall.req.mkdir.attributes, + ORANGEFS_TYPE_DIRECTORY, mode); + + strncpy(new_op->upcall.req.mkdir.d_name, + dentry->d_name.name, ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, __func__, get_interruptible_flag(dir)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Mkdir Got ORANGEFS handle %pU on fsid %d\n", + &new_op->downcall.resp.mkdir.refn.khandle, + new_op->downcall.resp.mkdir.refn.fs_id); + + if (ret < 0) { + gossip_debug(GOSSIP_NAME_DEBUG, + "%s: failed with error code %d\n", + __func__, ret); + goto out; + } + + inode = orangefs_new_inode(dir->i_sb, dir, S_IFDIR | mode, 0, + &new_op->downcall.resp.mkdir.refn); + if (IS_ERR(inode)) { + gossip_err("*** Failed to allocate orangefs dir inode\n"); + ret = PTR_ERR(inode); + goto out; + } + + gossip_debug(GOSSIP_NAME_DEBUG, + "Assigned dir inode new number of %pU\n", + get_khandle_from_ino(inode)); + + d_instantiate(dentry, inode); + unlock_new_inode(inode); + + gossip_debug(GOSSIP_NAME_DEBUG, + "Inode (Directory) %pU -> %s\n", + get_khandle_from_ino(inode), + dentry->d_name.name); + + /* + * NOTE: we have no good way to keep nlink consistent for directories + * across clients; keep constant at 1. + */ + SetMtimeFlag(parent); + dir->i_mtime = dir->i_ctime = current_fs_time(dir->i_sb); + mark_inode_dirty_sync(dir); +out: + op_release(new_op); + return ret; +} + +static int orangefs_rename(struct inode *old_dir, + struct dentry *old_dentry, + struct inode *new_dir, + struct dentry *new_dentry) +{ + struct orangefs_kernel_op_s *new_op; + int ret; + + gossip_debug(GOSSIP_NAME_DEBUG, + "orangefs_rename: called (%s/%s => %s/%s) ct=%d\n", + old_dentry->d_parent->d_name.name, + old_dentry->d_name.name, + new_dentry->d_parent->d_name.name, + new_dentry->d_name.name, + d_count(new_dentry)); + + new_op = op_alloc(ORANGEFS_VFS_OP_RENAME); + if (!new_op) + return -EINVAL; + + new_op->upcall.req.rename.old_parent_refn = ORANGEFS_I(old_dir)->refn; + new_op->upcall.req.rename.new_parent_refn = ORANGEFS_I(new_dir)->refn; + + strncpy(new_op->upcall.req.rename.d_old_name, + old_dentry->d_name.name, + ORANGEFS_NAME_MAX); + strncpy(new_op->upcall.req.rename.d_new_name, + new_dentry->d_name.name, + ORANGEFS_NAME_MAX); + + ret = service_operation(new_op, + "orangefs_rename", + get_interruptible_flag(old_dentry->d_inode)); + + gossip_debug(GOSSIP_NAME_DEBUG, + "orangefs_rename: got downcall status %d\n", + ret); + + if (new_dentry->d_inode) + new_dentry->d_inode->i_ctime = CURRENT_TIME; + + op_release(new_op); + return ret; +} + +/* ORANGEFS implementation of VFS inode operations for directories */ +struct inode_operations orangefs_dir_inode_operations = { + .lookup = orangefs_lookup, + .get_acl = orangefs_get_acl, + .set_acl = orangefs_set_acl, + .create = orangefs_create, + .unlink = orangefs_unlink, + .symlink = orangefs_symlink, + .mkdir = orangefs_mkdir, + .rmdir = orangefs_unlink, + .rename = orangefs_rename, + .setattr = orangefs_setattr, + .getattr = orangefs_getattr, + .setxattr = generic_setxattr, + .getxattr = generic_getxattr, + .removexattr = generic_removexattr, + .listxattr = orangefs_listxattr, + .permission = orangefs_permission, +}; diff --git a/fs/orangefs/orangefs-bufmap.c b/fs/orangefs/orangefs-bufmap.c new file mode 100644 index 000000000000..1f8acc9f9a88 --- /dev/null +++ b/fs/orangefs/orangefs-bufmap.c @@ -0,0 +1,556 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +struct slot_map { + int c; + wait_queue_head_t q; + int count; + unsigned long *map; +}; + +static struct slot_map rw_map = { + .c = -1, + .q = __WAIT_QUEUE_HEAD_INITIALIZER(rw_map.q) +}; +static struct slot_map readdir_map = { + .c = -1, + .q = __WAIT_QUEUE_HEAD_INITIALIZER(readdir_map.q) +}; + + +static void install(struct slot_map *m, int count, unsigned long *map) +{ + spin_lock(&m->q.lock); + m->c = m->count = count; + m->map = map; + wake_up_all_locked(&m->q); + spin_unlock(&m->q.lock); +} + +static void mark_killed(struct slot_map *m) +{ + spin_lock(&m->q.lock); + m->c -= m->count + 1; + spin_unlock(&m->q.lock); +} + +static void run_down(struct slot_map *m) +{ + DEFINE_WAIT(wait); + spin_lock(&m->q.lock); + if (m->c != -1) { + for (;;) { + if (likely(list_empty(&wait.task_list))) + __add_wait_queue_tail(&m->q, &wait); + set_current_state(TASK_UNINTERRUPTIBLE); + + if (m->c == -1) + break; + + spin_unlock(&m->q.lock); + schedule(); + spin_lock(&m->q.lock); + } + __remove_wait_queue(&m->q, &wait); + __set_current_state(TASK_RUNNING); + } + m->map = NULL; + spin_unlock(&m->q.lock); +} + +static void put(struct slot_map *m, int slot) +{ + int v; + spin_lock(&m->q.lock); + __clear_bit(slot, m->map); + v = ++m->c; + if (unlikely(v == 1)) /* no free slots -> one free slot */ + wake_up_locked(&m->q); + else if (unlikely(v == -1)) /* finished dying */ + wake_up_all_locked(&m->q); + spin_unlock(&m->q.lock); +} + +static int wait_for_free(struct slot_map *m) +{ + long left = slot_timeout_secs * HZ; + DEFINE_WAIT(wait); + + do { + long n = left, t; + if (likely(list_empty(&wait.task_list))) + __add_wait_queue_tail_exclusive(&m->q, &wait); + set_current_state(TASK_INTERRUPTIBLE); + + if (m->c > 0) + break; + + if (m->c < 0) { + /* we are waiting for map to be installed */ + /* it would better be there soon, or we go away */ + if (n > ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ) + n = ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS * HZ; + } + spin_unlock(&m->q.lock); + t = schedule_timeout(n); + spin_lock(&m->q.lock); + if (unlikely(!t) && n != left && m->c < 0) + left = t; + else + left = t + (left - n); + if (unlikely(signal_pending(current))) + left = -EINTR; + } while (left > 0); + + if (!list_empty(&wait.task_list)) + list_del(&wait.task_list); + else if (left <= 0 && waitqueue_active(&m->q)) + __wake_up_locked_key(&m->q, TASK_INTERRUPTIBLE, NULL); + __set_current_state(TASK_RUNNING); + + if (likely(left > 0)) + return 0; + + return left < 0 ? -EINTR : -ETIMEDOUT; +} + +static int get(struct slot_map *m) +{ + int res = 0; + spin_lock(&m->q.lock); + if (unlikely(m->c <= 0)) + res = wait_for_free(m); + if (likely(!res)) { + m->c--; + res = find_first_zero_bit(m->map, m->count); + __set_bit(res, m->map); + } + spin_unlock(&m->q.lock); + return res; +} + +/* used to describe mapped buffers */ +struct orangefs_bufmap_desc { + void *uaddr; /* user space address pointer */ + struct page **page_array; /* array of mapped pages */ + int array_count; /* size of above arrays */ + struct list_head list_link; +}; + +static struct orangefs_bufmap { + int desc_size; + int desc_shift; + int desc_count; + int total_size; + int page_count; + + struct page **page_array; + struct orangefs_bufmap_desc *desc_array; + + /* array to track usage of buffer descriptors */ + unsigned long *buffer_index_array; + + /* array to track usage of buffer descriptors for readdir */ +#define N DIV_ROUND_UP(ORANGEFS_READDIR_DEFAULT_DESC_COUNT, BITS_PER_LONG) + unsigned long readdir_index_array[N]; +#undef N +} *__orangefs_bufmap; + +static DEFINE_SPINLOCK(orangefs_bufmap_lock); + +static void +orangefs_bufmap_unmap(struct orangefs_bufmap *bufmap) +{ + int i; + + for (i = 0; i < bufmap->page_count; i++) + page_cache_release(bufmap->page_array[i]); +} + +static void +orangefs_bufmap_free(struct orangefs_bufmap *bufmap) +{ + kfree(bufmap->page_array); + kfree(bufmap->desc_array); + kfree(bufmap->buffer_index_array); + kfree(bufmap); +} + +/* + * XXX: Can the size and shift change while the caller gives up the + * XXX: lock between calling this and doing something useful? + */ + +int orangefs_bufmap_size_query(void) +{ + struct orangefs_bufmap *bufmap; + int size = 0; + spin_lock(&orangefs_bufmap_lock); + bufmap = __orangefs_bufmap; + if (bufmap) + size = bufmap->desc_size; + spin_unlock(&orangefs_bufmap_lock); + return size; +} + +int orangefs_bufmap_shift_query(void) +{ + struct orangefs_bufmap *bufmap; + int shift = 0; + spin_lock(&orangefs_bufmap_lock); + bufmap = __orangefs_bufmap; + if (bufmap) + shift = bufmap->desc_shift; + spin_unlock(&orangefs_bufmap_lock); + return shift; +} + +static DECLARE_WAIT_QUEUE_HEAD(bufmap_waitq); +static DECLARE_WAIT_QUEUE_HEAD(readdir_waitq); + +/* + * orangefs_get_bufmap_init + * + * If bufmap_init is 1, then the shared memory system, including the + * buffer_index_array, is available. Otherwise, it is not. + * + * returns the value of bufmap_init + */ +int orangefs_get_bufmap_init(void) +{ + return __orangefs_bufmap ? 1 : 0; +} + + +static struct orangefs_bufmap * +orangefs_bufmap_alloc(struct ORANGEFS_dev_map_desc *user_desc) +{ + struct orangefs_bufmap *bufmap; + + bufmap = kzalloc(sizeof(*bufmap), GFP_KERNEL); + if (!bufmap) + goto out; + + bufmap->total_size = user_desc->total_size; + bufmap->desc_count = user_desc->count; + bufmap->desc_size = user_desc->size; + bufmap->desc_shift = ilog2(bufmap->desc_size); + + bufmap->buffer_index_array = + kzalloc(DIV_ROUND_UP(bufmap->desc_count, BITS_PER_LONG), GFP_KERNEL); + if (!bufmap->buffer_index_array) { + gossip_err("orangefs: could not allocate %d buffer indices\n", + bufmap->desc_count); + goto out_free_bufmap; + } + + bufmap->desc_array = + kcalloc(bufmap->desc_count, sizeof(struct orangefs_bufmap_desc), + GFP_KERNEL); + if (!bufmap->desc_array) { + gossip_err("orangefs: could not allocate %d descriptors\n", + bufmap->desc_count); + goto out_free_index_array; + } + + bufmap->page_count = bufmap->total_size / PAGE_SIZE; + + /* allocate storage to track our page mappings */ + bufmap->page_array = + kcalloc(bufmap->page_count, sizeof(struct page *), GFP_KERNEL); + if (!bufmap->page_array) + goto out_free_desc_array; + + return bufmap; + +out_free_desc_array: + kfree(bufmap->desc_array); +out_free_index_array: + kfree(bufmap->buffer_index_array); +out_free_bufmap: + kfree(bufmap); +out: + return NULL; +} + +static int +orangefs_bufmap_map(struct orangefs_bufmap *bufmap, + struct ORANGEFS_dev_map_desc *user_desc) +{ + int pages_per_desc = bufmap->desc_size / PAGE_SIZE; + int offset = 0, ret, i; + + /* map the pages */ + ret = get_user_pages_fast((unsigned long)user_desc->ptr, + bufmap->page_count, 1, bufmap->page_array); + + if (ret < 0) + return ret; + + if (ret != bufmap->page_count) { + gossip_err("orangefs error: asked for %d pages, only got %d.\n", + bufmap->page_count, ret); + + for (i = 0; i < ret; i++) { + SetPageError(bufmap->page_array[i]); + page_cache_release(bufmap->page_array[i]); + } + return -ENOMEM; + } + + /* + * ideally we want to get kernel space pointers for each page, but + * we can't kmap that many pages at once if highmem is being used. + * so instead, we just kmap/kunmap the page address each time the + * kaddr is needed. + */ + for (i = 0; i < bufmap->page_count; i++) + flush_dcache_page(bufmap->page_array[i]); + + /* build a list of available descriptors */ + for (offset = 0, i = 0; i < bufmap->desc_count; i++) { + bufmap->desc_array[i].page_array = &bufmap->page_array[offset]; + bufmap->desc_array[i].array_count = pages_per_desc; + bufmap->desc_array[i].uaddr = + (user_desc->ptr + (i * pages_per_desc * PAGE_SIZE)); + offset += pages_per_desc; + } + + return 0; +} + +/* + * orangefs_bufmap_initialize() + * + * initializes the mapped buffer interface + * + * returns 0 on success, -errno on failure + */ +int orangefs_bufmap_initialize(struct ORANGEFS_dev_map_desc *user_desc) +{ + struct orangefs_bufmap *bufmap; + int ret = -EINVAL; + + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "orangefs_bufmap_initialize: called (ptr (" + "%p) sz (%d) cnt(%d).\n", + user_desc->ptr, + user_desc->size, + user_desc->count); + + /* + * sanity check alignment and size of buffer that caller wants to + * work with + */ + if (PAGE_ALIGN((unsigned long)user_desc->ptr) != + (unsigned long)user_desc->ptr) { + gossip_err("orangefs error: memory alignment (front). %p\n", + user_desc->ptr); + goto out; + } + + if (PAGE_ALIGN(((unsigned long)user_desc->ptr + user_desc->total_size)) + != (unsigned long)(user_desc->ptr + user_desc->total_size)) { + gossip_err("orangefs error: memory alignment (back).(%p + %d)\n", + user_desc->ptr, + user_desc->total_size); + goto out; + } + + if (user_desc->total_size != (user_desc->size * user_desc->count)) { + gossip_err("orangefs error: user provided an oddly sized buffer: (%d, %d, %d)\n", + user_desc->total_size, + user_desc->size, + user_desc->count); + goto out; + } + + if ((user_desc->size % PAGE_SIZE) != 0) { + gossip_err("orangefs error: bufmap size not page size divisible (%d).\n", + user_desc->size); + goto out; + } + + ret = -ENOMEM; + bufmap = orangefs_bufmap_alloc(user_desc); + if (!bufmap) + goto out; + + ret = orangefs_bufmap_map(bufmap, user_desc); + if (ret) + goto out_free_bufmap; + + + spin_lock(&orangefs_bufmap_lock); + if (__orangefs_bufmap) { + spin_unlock(&orangefs_bufmap_lock); + gossip_err("orangefs: error: bufmap already initialized.\n"); + ret = -EINVAL; + goto out_unmap_bufmap; + } + __orangefs_bufmap = bufmap; + install(&rw_map, + bufmap->desc_count, + bufmap->buffer_index_array); + install(&readdir_map, + ORANGEFS_READDIR_DEFAULT_DESC_COUNT, + bufmap->readdir_index_array); + spin_unlock(&orangefs_bufmap_lock); + + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "orangefs_bufmap_initialize: exiting normally\n"); + return 0; + +out_unmap_bufmap: + orangefs_bufmap_unmap(bufmap); +out_free_bufmap: + orangefs_bufmap_free(bufmap); +out: + return ret; +} + +/* + * orangefs_bufmap_finalize() + * + * shuts down the mapped buffer interface and releases any resources + * associated with it + * + * no return value + */ +void orangefs_bufmap_finalize(void) +{ + struct orangefs_bufmap *bufmap = __orangefs_bufmap; + if (!bufmap) + return; + gossip_debug(GOSSIP_BUFMAP_DEBUG, "orangefs_bufmap_finalize: called\n"); + mark_killed(&rw_map); + mark_killed(&readdir_map); + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "orangefs_bufmap_finalize: exiting normally\n"); +} + +void orangefs_bufmap_run_down(void) +{ + struct orangefs_bufmap *bufmap = __orangefs_bufmap; + if (!bufmap) + return; + run_down(&rw_map); + run_down(&readdir_map); + spin_lock(&orangefs_bufmap_lock); + __orangefs_bufmap = NULL; + spin_unlock(&orangefs_bufmap_lock); + orangefs_bufmap_unmap(bufmap); + orangefs_bufmap_free(bufmap); +} + +/* + * orangefs_bufmap_get() + * + * gets a free mapped buffer descriptor, will sleep until one becomes + * available if necessary + * + * returns slot on success, -errno on failure + */ +int orangefs_bufmap_get(void) +{ + return get(&rw_map); +} + +/* + * orangefs_bufmap_put() + * + * returns a mapped buffer descriptor to the collection + * + * no return value + */ +void orangefs_bufmap_put(int buffer_index) +{ + put(&rw_map, buffer_index); +} + +/* + * orangefs_readdir_index_get() + * + * gets a free descriptor, will sleep until one becomes + * available if necessary. + * Although the readdir buffers are not mapped into kernel space + * we could do that at a later point of time. Regardless, these + * indices are used by the client-core. + * + * returns slot on success, -errno on failure + */ +int orangefs_readdir_index_get(void) +{ + return get(&readdir_map); +} + +void orangefs_readdir_index_put(int buffer_index) +{ + put(&readdir_map, buffer_index); +} + +/* + * we've been handed an iovec, we need to copy it to + * the shared memory descriptor at "buffer_index". + */ +int orangefs_bufmap_copy_from_iovec(struct iov_iter *iter, + int buffer_index, + size_t size) +{ + struct orangefs_bufmap_desc *to; + int i; + + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "%s: buffer_index:%d: size:%zu:\n", + __func__, buffer_index, size); + + to = &__orangefs_bufmap->desc_array[buffer_index]; + for (i = 0; size; i++) { + struct page *page = to->page_array[i]; + size_t n = size; + if (n > PAGE_SIZE) + n = PAGE_SIZE; + n = copy_page_from_iter(page, 0, n, iter); + if (!n) + return -EFAULT; + size -= n; + } + return 0; + +} + +/* + * we've been handed an iovec, we need to fill it from + * the shared memory descriptor at "buffer_index". + */ +int orangefs_bufmap_copy_to_iovec(struct iov_iter *iter, + int buffer_index, + size_t size) +{ + struct orangefs_bufmap_desc *from; + int i; + + from = &__orangefs_bufmap->desc_array[buffer_index]; + gossip_debug(GOSSIP_BUFMAP_DEBUG, + "%s: buffer_index:%d: size:%zu:\n", + __func__, buffer_index, size); + + + for (i = 0; size; i++) { + struct page *page = from->page_array[i]; + size_t n = size; + if (n > PAGE_SIZE) + n = PAGE_SIZE; + n = copy_page_to_iter(page, 0, n, iter); + if (!n) + return -EFAULT; + size -= n; + } + return 0; +} diff --git a/fs/orangefs/orangefs-bufmap.h b/fs/orangefs/orangefs-bufmap.h new file mode 100644 index 000000000000..71f64f4057b5 --- /dev/null +++ b/fs/orangefs/orangefs-bufmap.h @@ -0,0 +1,36 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#ifndef __ORANGEFS_BUFMAP_H +#define __ORANGEFS_BUFMAP_H + +int orangefs_bufmap_size_query(void); + +int orangefs_bufmap_shift_query(void); + +int orangefs_bufmap_initialize(struct ORANGEFS_dev_map_desc *user_desc); + +void orangefs_bufmap_finalize(void); + +void orangefs_bufmap_run_down(void); + +int orangefs_bufmap_get(void); + +void orangefs_bufmap_put(int buffer_index); + +int orangefs_readdir_index_get(void); + +void orangefs_readdir_index_put(int buffer_index); + +int orangefs_bufmap_copy_from_iovec(struct iov_iter *iter, + int buffer_index, + size_t size); + +int orangefs_bufmap_copy_to_iovec(struct iov_iter *iter, + int buffer_index, + size_t size); + +#endif /* __ORANGEFS_BUFMAP_H */ diff --git a/fs/orangefs/orangefs-cache.c b/fs/orangefs/orangefs-cache.c new file mode 100644 index 000000000000..900a2e38e11b --- /dev/null +++ b/fs/orangefs/orangefs-cache.c @@ -0,0 +1,161 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" + +/* tags assigned to kernel upcall operations */ +static __u64 next_tag_value; +static DEFINE_SPINLOCK(next_tag_value_lock); + +/* the orangefs memory caches */ + +/* a cache for orangefs upcall/downcall operations */ +static struct kmem_cache *op_cache; + +int op_cache_initialize(void) +{ + op_cache = kmem_cache_create("orangefs_op_cache", + sizeof(struct orangefs_kernel_op_s), + 0, + ORANGEFS_CACHE_CREATE_FLAGS, + NULL); + + if (!op_cache) { + gossip_err("Cannot create orangefs_op_cache\n"); + return -ENOMEM; + } + + /* initialize our atomic tag counter */ + spin_lock(&next_tag_value_lock); + next_tag_value = 100; + spin_unlock(&next_tag_value_lock); + return 0; +} + +int op_cache_finalize(void) +{ + kmem_cache_destroy(op_cache); + return 0; +} + +char *get_opname_string(struct orangefs_kernel_op_s *new_op) +{ + if (new_op) { + __s32 type = new_op->upcall.type; + + if (type == ORANGEFS_VFS_OP_FILE_IO) + return "OP_FILE_IO"; + else if (type == ORANGEFS_VFS_OP_LOOKUP) + return "OP_LOOKUP"; + else if (type == ORANGEFS_VFS_OP_CREATE) + return "OP_CREATE"; + else if (type == ORANGEFS_VFS_OP_GETATTR) + return "OP_GETATTR"; + else if (type == ORANGEFS_VFS_OP_REMOVE) + return "OP_REMOVE"; + else if (type == ORANGEFS_VFS_OP_MKDIR) + return "OP_MKDIR"; + else if (type == ORANGEFS_VFS_OP_READDIR) + return "OP_READDIR"; + else if (type == ORANGEFS_VFS_OP_READDIRPLUS) + return "OP_READDIRPLUS"; + else if (type == ORANGEFS_VFS_OP_SETATTR) + return "OP_SETATTR"; + else if (type == ORANGEFS_VFS_OP_SYMLINK) + return "OP_SYMLINK"; + else if (type == ORANGEFS_VFS_OP_RENAME) + return "OP_RENAME"; + else if (type == ORANGEFS_VFS_OP_STATFS) + return "OP_STATFS"; + else if (type == ORANGEFS_VFS_OP_TRUNCATE) + return "OP_TRUNCATE"; + else if (type == ORANGEFS_VFS_OP_MMAP_RA_FLUSH) + return "OP_MMAP_RA_FLUSH"; + else if (type == ORANGEFS_VFS_OP_FS_MOUNT) + return "OP_FS_MOUNT"; + else if (type == ORANGEFS_VFS_OP_FS_UMOUNT) + return "OP_FS_UMOUNT"; + else if (type == ORANGEFS_VFS_OP_GETXATTR) + return "OP_GETXATTR"; + else if (type == ORANGEFS_VFS_OP_SETXATTR) + return "OP_SETXATTR"; + else if (type == ORANGEFS_VFS_OP_LISTXATTR) + return "OP_LISTXATTR"; + else if (type == ORANGEFS_VFS_OP_REMOVEXATTR) + return "OP_REMOVEXATTR"; + else if (type == ORANGEFS_VFS_OP_PARAM) + return "OP_PARAM"; + else if (type == ORANGEFS_VFS_OP_PERF_COUNT) + return "OP_PERF_COUNT"; + else if (type == ORANGEFS_VFS_OP_CANCEL) + return "OP_CANCEL"; + else if (type == ORANGEFS_VFS_OP_FSYNC) + return "OP_FSYNC"; + else if (type == ORANGEFS_VFS_OP_FSKEY) + return "OP_FSKEY"; + } + return "OP_UNKNOWN?"; +} + +void orangefs_new_tag(struct orangefs_kernel_op_s *op) +{ + spin_lock(&next_tag_value_lock); + op->tag = next_tag_value++; + if (next_tag_value == 0) + next_tag_value = 100; + spin_unlock(&next_tag_value_lock); +} + +struct orangefs_kernel_op_s *op_alloc(__s32 type) +{ + struct orangefs_kernel_op_s *new_op = NULL; + + new_op = kmem_cache_zalloc(op_cache, GFP_KERNEL); + if (new_op) { + INIT_LIST_HEAD(&new_op->list); + spin_lock_init(&new_op->lock); + init_completion(&new_op->waitq); + + new_op->upcall.type = ORANGEFS_VFS_OP_INVALID; + new_op->downcall.type = ORANGEFS_VFS_OP_INVALID; + new_op->downcall.status = -1; + + new_op->op_state = OP_VFS_STATE_UNKNOWN; + + /* initialize the op specific tag and upcall credentials */ + orangefs_new_tag(new_op); + new_op->upcall.type = type; + new_op->attempts = 0; + gossip_debug(GOSSIP_CACHE_DEBUG, + "Alloced OP (%p: %llu %s)\n", + new_op, + llu(new_op->tag), + get_opname_string(new_op)); + + new_op->upcall.uid = from_kuid(current_user_ns(), + current_fsuid()); + + new_op->upcall.gid = from_kgid(current_user_ns(), + current_fsgid()); + } else { + gossip_err("op_alloc: kmem_cache_zalloc failed!\n"); + } + return new_op; +} + +void op_release(struct orangefs_kernel_op_s *orangefs_op) +{ + if (orangefs_op) { + gossip_debug(GOSSIP_CACHE_DEBUG, + "Releasing OP (%p: %llu)\n", + orangefs_op, + llu(orangefs_op->tag)); + kmem_cache_free(op_cache, orangefs_op); + } else { + gossip_err("NULL pointer in op_release\n"); + } +} diff --git a/fs/orangefs/orangefs-debug.h b/fs/orangefs/orangefs-debug.h new file mode 100644 index 000000000000..387db17cde2b --- /dev/null +++ b/fs/orangefs/orangefs-debug.h @@ -0,0 +1,92 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* This file just defines debugging masks to be used with the gossip + * logging utility. All debugging masks for ORANGEFS are kept here to make + * sure we don't have collisions. + */ + +#ifndef __ORANGEFS_DEBUG_H +#define __ORANGEFS_DEBUG_H + +#ifdef __KERNEL__ +#include <linux/types.h> +#else +#include <stdint.h> +#endif + +#define GOSSIP_NO_DEBUG (__u64)0 + +#define GOSSIP_SUPER_DEBUG ((__u64)1 << 0) +#define GOSSIP_INODE_DEBUG ((__u64)1 << 1) +#define GOSSIP_FILE_DEBUG ((__u64)1 << 2) +#define GOSSIP_DIR_DEBUG ((__u64)1 << 3) +#define GOSSIP_UTILS_DEBUG ((__u64)1 << 4) +#define GOSSIP_WAIT_DEBUG ((__u64)1 << 5) +#define GOSSIP_ACL_DEBUG ((__u64)1 << 6) +#define GOSSIP_DCACHE_DEBUG ((__u64)1 << 7) +#define GOSSIP_DEV_DEBUG ((__u64)1 << 8) +#define GOSSIP_NAME_DEBUG ((__u64)1 << 9) +#define GOSSIP_BUFMAP_DEBUG ((__u64)1 << 10) +#define GOSSIP_CACHE_DEBUG ((__u64)1 << 11) +#define GOSSIP_DEBUGFS_DEBUG ((__u64)1 << 12) +#define GOSSIP_XATTR_DEBUG ((__u64)1 << 13) +#define GOSSIP_INIT_DEBUG ((__u64)1 << 14) +#define GOSSIP_SYSFS_DEBUG ((__u64)1 << 15) + +#define GOSSIP_MAX_NR 16 +#define GOSSIP_MAX_DEBUG (((__u64)1 << GOSSIP_MAX_NR) - 1) + +/*function prototypes*/ +__u64 ORANGEFS_kmod_eventlog_to_mask(const char *event_logging); +__u64 ORANGEFS_debug_eventlog_to_mask(const char *event_logging); +char *ORANGEFS_debug_mask_to_eventlog(__u64 mask); +char *ORANGEFS_kmod_mask_to_eventlog(__u64 mask); + +/* a private internal type */ +struct __keyword_mask_s { + const char *keyword; + __u64 mask_val; +}; + +/* + * Map all kmod keywords to kmod debug masks here. Keep this + * structure "packed": + * + * "all" is always last... + * + * keyword mask_val index + * foo 1 0 + * bar 2 1 + * baz 4 2 + * qux 8 3 + * . . . + */ +static struct __keyword_mask_s s_kmod_keyword_mask_map[] = { + {"super", GOSSIP_SUPER_DEBUG}, + {"inode", GOSSIP_INODE_DEBUG}, + {"file", GOSSIP_FILE_DEBUG}, + {"dir", GOSSIP_DIR_DEBUG}, + {"utils", GOSSIP_UTILS_DEBUG}, + {"wait", GOSSIP_WAIT_DEBUG}, + {"acl", GOSSIP_ACL_DEBUG}, + {"dcache", GOSSIP_DCACHE_DEBUG}, + {"dev", GOSSIP_DEV_DEBUG}, + {"name", GOSSIP_NAME_DEBUG}, + {"bufmap", GOSSIP_BUFMAP_DEBUG}, + {"cache", GOSSIP_CACHE_DEBUG}, + {"debugfs", GOSSIP_DEBUGFS_DEBUG}, + {"xattr", GOSSIP_XATTR_DEBUG}, + {"init", GOSSIP_INIT_DEBUG}, + {"sysfs", GOSSIP_SYSFS_DEBUG}, + {"none", GOSSIP_NO_DEBUG}, + {"all", GOSSIP_MAX_DEBUG} +}; + +static const int num_kmod_keyword_mask_map = (int) + (sizeof(s_kmod_keyword_mask_map) / sizeof(struct __keyword_mask_s)); + +#endif /* __ORANGEFS_DEBUG_H */ diff --git a/fs/orangefs/orangefs-debugfs.c b/fs/orangefs/orangefs-debugfs.c new file mode 100644 index 000000000000..19670b8b4053 --- /dev/null +++ b/fs/orangefs/orangefs-debugfs.c @@ -0,0 +1,455 @@ +/* + * What: /sys/kernel/debug/orangefs/debug-help + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * List of client and kernel debug keywords. + * + * + * What: /sys/kernel/debug/orangefs/client-debug + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Debug setting for "the client", the userspace + * helper for the kernel module. + * + * + * What: /sys/kernel/debug/orangefs/kernel-debug + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Debug setting for the orangefs kernel module. + * + * Any of the keywords, or comma-separated lists + * of keywords, from debug-help can be catted to + * client-debug or kernel-debug. + * + * "none", "all" and "verbose" are special keywords + * for client-debug. Setting client-debug to "all" + * is kind of like trying to drink water from a + * fire hose, "verbose" triggers most of the same + * output except for the constant flow of output + * from the main wait loop. + * + * "none" and "all" are similar settings for kernel-debug + * no need for a "verbose". + */ +#include <linux/debugfs.h> +#include <linux/slab.h> + +#include <linux/uaccess.h> + +#include "orangefs-debugfs.h" +#include "protocol.h" +#include "orangefs-kernel.h" + +static int orangefs_debug_disabled = 1; + +static int orangefs_debug_help_open(struct inode *, struct file *); + +const struct file_operations debug_help_fops = { + .open = orangefs_debug_help_open, + .read = seq_read, + .release = seq_release, + .llseek = seq_lseek, +}; + +static void *help_start(struct seq_file *, loff_t *); +static void *help_next(struct seq_file *, void *, loff_t *); +static void help_stop(struct seq_file *, void *); +static int help_show(struct seq_file *, void *); + +static const struct seq_operations help_debug_ops = { + .start = help_start, + .next = help_next, + .stop = help_stop, + .show = help_show, +}; + +/* + * Used to protect data in ORANGEFS_KMOD_DEBUG_FILE and + * ORANGEFS_KMOD_DEBUG_FILE. + */ +static DEFINE_MUTEX(orangefs_debug_lock); + +int orangefs_debug_open(struct inode *, struct file *); + +static ssize_t orangefs_debug_read(struct file *, + char __user *, + size_t, + loff_t *); + +static ssize_t orangefs_debug_write(struct file *, + const char __user *, + size_t, + loff_t *); + +static const struct file_operations kernel_debug_fops = { + .open = orangefs_debug_open, + .read = orangefs_debug_read, + .write = orangefs_debug_write, + .llseek = generic_file_llseek, +}; + +/* + * initialize kmod debug operations, create orangefs debugfs dir and + * ORANGEFS_KMOD_DEBUG_HELP_FILE. + */ +int orangefs_debugfs_init(void) +{ + + int rc = -ENOMEM; + + debug_dir = debugfs_create_dir("orangefs", NULL); + if (!debug_dir) { + pr_info("%s: debugfs_create_dir failed.\n", __func__); + goto out; + } + + help_file_dentry = debugfs_create_file(ORANGEFS_KMOD_DEBUG_HELP_FILE, + 0444, + debug_dir, + debug_help_string, + &debug_help_fops); + if (!help_file_dentry) { + pr_info("%s: debugfs_create_file failed.\n", __func__); + goto out; + } + + orangefs_debug_disabled = 0; + rc = 0; + +out: + + return rc; +} + +void orangefs_debugfs_cleanup(void) +{ + if (debug_dir) + debugfs_remove_recursive(debug_dir); +} + +/* open ORANGEFS_KMOD_DEBUG_HELP_FILE */ +static int orangefs_debug_help_open(struct inode *inode, struct file *file) +{ + int rc = -ENODEV; + int ret; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_help_open: start\n"); + + if (orangefs_debug_disabled) + goto out; + + ret = seq_open(file, &help_debug_ops); + if (ret) + goto out; + + ((struct seq_file *)(file->private_data))->private = inode->i_private; + + rc = 0; + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_help_open: rc:%d:\n", + rc); + return rc; +} + +/* + * I think start always gets called again after stop. Start + * needs to return NULL when it is done. The whole "payload" + * in this case is a single (long) string, so by the second + * time we get to start (pos = 1), we're done. + */ +static void *help_start(struct seq_file *m, loff_t *pos) +{ + void *payload = NULL; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_start: start\n"); + + if (*pos == 0) + payload = m->private; + + return payload; +} + +static void *help_next(struct seq_file *m, void *v, loff_t *pos) +{ + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_next: start\n"); + + return NULL; +} + +static void help_stop(struct seq_file *m, void *p) +{ + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_stop: start\n"); +} + +static int help_show(struct seq_file *m, void *v) +{ + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "help_show: start\n"); + + seq_puts(m, v); + + return 0; +} + +/* + * initialize the kernel-debug file. + */ +int orangefs_kernel_debug_init(void) +{ + int rc = -ENOMEM; + struct dentry *ret; + char *k_buffer = NULL; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: start\n", __func__); + + k_buffer = kzalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!k_buffer) + goto out; + + if (strlen(kernel_debug_string) + 1 < ORANGEFS_MAX_DEBUG_STRING_LEN) { + strcpy(k_buffer, kernel_debug_string); + strcat(k_buffer, "\n"); + } else { + strcpy(k_buffer, "none\n"); + pr_info("%s: overflow 1!\n", __func__); + } + + ret = debugfs_create_file(ORANGEFS_KMOD_DEBUG_FILE, + 0444, + debug_dir, + k_buffer, + &kernel_debug_fops); + if (!ret) { + pr_info("%s: failed to create %s.\n", + __func__, + ORANGEFS_KMOD_DEBUG_FILE); + goto out; + } + + rc = 0; + +out: + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: rc:%d:\n", __func__, rc); + return rc; +} + +/* + * initialize the client-debug file. + */ +int orangefs_client_debug_init(void) +{ + + int rc = -ENOMEM; + char *c_buffer = NULL; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: start\n", __func__); + + c_buffer = kzalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!c_buffer) + goto out; + + if (strlen(client_debug_string) + 1 < ORANGEFS_MAX_DEBUG_STRING_LEN) { + strcpy(c_buffer, client_debug_string); + strcat(c_buffer, "\n"); + } else { + strcpy(c_buffer, "none\n"); + pr_info("%s: overflow! 2\n", __func__); + } + + client_debug_dentry = debugfs_create_file(ORANGEFS_CLIENT_DEBUG_FILE, + 0444, + debug_dir, + c_buffer, + &kernel_debug_fops); + if (!client_debug_dentry) { + pr_info("%s: failed to create updated %s.\n", + __func__, + ORANGEFS_CLIENT_DEBUG_FILE); + goto out; + } + + rc = 0; + +out: + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "%s: rc:%d:\n", __func__, rc); + return rc; +} + +/* open ORANGEFS_KMOD_DEBUG_FILE or ORANGEFS_CLIENT_DEBUG_FILE.*/ +int orangefs_debug_open(struct inode *inode, struct file *file) +{ + int rc = -ENODEV; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "%s: orangefs_debug_disabled: %d\n", + __func__, + orangefs_debug_disabled); + + if (orangefs_debug_disabled) + goto out; + + rc = 0; + mutex_lock(&orangefs_debug_lock); + file->private_data = inode->i_private; + mutex_unlock(&orangefs_debug_lock); + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_open: rc: %d\n", + rc); + return rc; +} + +static ssize_t orangefs_debug_read(struct file *file, + char __user *ubuf, + size_t count, + loff_t *ppos) +{ + char *buf; + int sprintf_ret; + ssize_t read_ret = -ENOMEM; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, "orangefs_debug_read: start\n"); + + buf = kmalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!buf) + goto out; + + mutex_lock(&orangefs_debug_lock); + sprintf_ret = sprintf(buf, "%s", (char *)file->private_data); + mutex_unlock(&orangefs_debug_lock); + + read_ret = simple_read_from_buffer(ubuf, count, ppos, buf, sprintf_ret); + + kfree(buf); + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_read: ret: %zu\n", + read_ret); + + return read_ret; +} + +static ssize_t orangefs_debug_write(struct file *file, + const char __user *ubuf, + size_t count, + loff_t *ppos) +{ + char *buf; + int rc = -EFAULT; + size_t silly = 0; + char *debug_string; + struct orangefs_kernel_op_s *new_op = NULL; + struct client_debug_mask c_mask = { NULL, 0, 0 }; + + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_write: %s\n", + file->f_path.dentry->d_name.name); + + /* + * Thwart users who try to jamb a ridiculous number + * of bytes into the debug file... + */ + if (count > ORANGEFS_MAX_DEBUG_STRING_LEN + 1) { + silly = count; + count = ORANGEFS_MAX_DEBUG_STRING_LEN + 1; + } + + buf = kzalloc(ORANGEFS_MAX_DEBUG_STRING_LEN, GFP_KERNEL); + if (!buf) + goto out; + + if (copy_from_user(buf, ubuf, count - 1)) { + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "%s: copy_from_user failed!\n", + __func__); + goto out; + } + + /* + * Map the keyword string from userspace into a valid debug mask. + * The mapping process involves mapping the human-inputted string + * into a valid mask, and then rebuilding the string from the + * verified valid mask. + * + * A service operation is required to set a new client-side + * debug mask. + */ + if (!strcmp(file->f_path.dentry->d_name.name, + ORANGEFS_KMOD_DEBUG_FILE)) { + debug_string_to_mask(buf, &gossip_debug_mask, 0); + debug_mask_to_string(&gossip_debug_mask, 0); + debug_string = kernel_debug_string; + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "New kernel debug string is %s\n", + kernel_debug_string); + } else { + /* Can't reset client debug mask if client is not running. */ + if (is_daemon_in_service()) { + pr_info("%s: Client not running :%d:\n", + __func__, + is_daemon_in_service()); + goto out; + } + + debug_string_to_mask(buf, &c_mask, 1); + debug_mask_to_string(&c_mask, 1); + debug_string = client_debug_string; + + new_op = op_alloc(ORANGEFS_VFS_OP_PARAM); + if (!new_op) { + pr_info("%s: op_alloc failed!\n", __func__); + goto out; + } + + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_TWO_MASK_VALUES; + new_op->upcall.req.param.type = ORANGEFS_PARAM_REQUEST_SET; + memset(new_op->upcall.req.param.s_value, + 0, + ORANGEFS_MAX_DEBUG_STRING_LEN); + sprintf(new_op->upcall.req.param.s_value, + "%llx %llx\n", + c_mask.mask1, + c_mask.mask2); + + /* service_operation returns 0 on success... */ + rc = service_operation(new_op, + "orangefs_param", + ORANGEFS_OP_INTERRUPTIBLE); + + if (rc) + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "%s: service_operation failed! rc:%d:\n", + __func__, + rc); + + op_release(new_op); + } + + mutex_lock(&orangefs_debug_lock); + memset(file->f_inode->i_private, 0, ORANGEFS_MAX_DEBUG_STRING_LEN); + sprintf((char *)file->f_inode->i_private, "%s\n", debug_string); + mutex_unlock(&orangefs_debug_lock); + + *ppos += count; + if (silly) + rc = silly; + else + rc = count; + +out: + gossip_debug(GOSSIP_DEBUGFS_DEBUG, + "orangefs_debug_write: rc: %d\n", + rc); + kfree(buf); + return rc; +} diff --git a/fs/orangefs/orangefs-debugfs.h b/fs/orangefs/orangefs-debugfs.h new file mode 100644 index 000000000000..e4828c0e3ef9 --- /dev/null +++ b/fs/orangefs/orangefs-debugfs.h @@ -0,0 +1,3 @@ +int orangefs_debugfs_init(void); +int orangefs_kernel_debug_init(void); +void orangefs_debugfs_cleanup(void); diff --git a/fs/orangefs/orangefs-dev-proto.h b/fs/orangefs/orangefs-dev-proto.h new file mode 100644 index 000000000000..9eac9d9a3f3a --- /dev/null +++ b/fs/orangefs/orangefs-dev-proto.h @@ -0,0 +1,62 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#ifndef _ORANGEFS_DEV_PROTO_H +#define _ORANGEFS_DEV_PROTO_H + +/* + * types and constants shared between user space and kernel space for + * device interaction using a common protocol + */ + +/* + * valid orangefs kernel operation types + */ +#define ORANGEFS_VFS_OP_INVALID 0xFF000000 +#define ORANGEFS_VFS_OP_FILE_IO 0xFF000001 +#define ORANGEFS_VFS_OP_LOOKUP 0xFF000002 +#define ORANGEFS_VFS_OP_CREATE 0xFF000003 +#define ORANGEFS_VFS_OP_GETATTR 0xFF000004 +#define ORANGEFS_VFS_OP_REMOVE 0xFF000005 +#define ORANGEFS_VFS_OP_MKDIR 0xFF000006 +#define ORANGEFS_VFS_OP_READDIR 0xFF000007 +#define ORANGEFS_VFS_OP_SETATTR 0xFF000008 +#define ORANGEFS_VFS_OP_SYMLINK 0xFF000009 +#define ORANGEFS_VFS_OP_RENAME 0xFF00000A +#define ORANGEFS_VFS_OP_STATFS 0xFF00000B +#define ORANGEFS_VFS_OP_TRUNCATE 0xFF00000C +#define ORANGEFS_VFS_OP_MMAP_RA_FLUSH 0xFF00000D +#define ORANGEFS_VFS_OP_FS_MOUNT 0xFF00000E +#define ORANGEFS_VFS_OP_FS_UMOUNT 0xFF00000F +#define ORANGEFS_VFS_OP_GETXATTR 0xFF000010 +#define ORANGEFS_VFS_OP_SETXATTR 0xFF000011 +#define ORANGEFS_VFS_OP_LISTXATTR 0xFF000012 +#define ORANGEFS_VFS_OP_REMOVEXATTR 0xFF000013 +#define ORANGEFS_VFS_OP_PARAM 0xFF000014 +#define ORANGEFS_VFS_OP_PERF_COUNT 0xFF000015 +#define ORANGEFS_VFS_OP_CANCEL 0xFF00EE00 +#define ORANGEFS_VFS_OP_FSYNC 0xFF00EE01 +#define ORANGEFS_VFS_OP_FSKEY 0xFF00EE02 +#define ORANGEFS_VFS_OP_READDIRPLUS 0xFF00EE03 + +/* + * Misc constants. Please retain them as multiples of 8! + * Otherwise 32-64 bit interactions will be messed up :) + */ +#define ORANGEFS_MAX_DEBUG_STRING_LEN 0x00000400 +#define ORANGEFS_MAX_DEBUG_ARRAY_LEN 0x00000800 + +/* + * The maximum number of directory entries in a single request is 96. + * XXX: Why can this not be higher. The client-side code can handle up to 512. + * XXX: What happens if we expect more than the client can return? + */ +#define ORANGEFS_MAX_DIRENT_COUNT_READDIR 96 + +#include "upcall.h" +#include "downcall.h" + +#endif diff --git a/fs/orangefs/orangefs-kernel.h b/fs/orangefs/orangefs-kernel.h new file mode 100644 index 000000000000..a9925e296ceb --- /dev/null +++ b/fs/orangefs/orangefs-kernel.h @@ -0,0 +1,623 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * The ORANGEFS Linux kernel support allows ORANGEFS volumes to be mounted and + * accessed through the Linux VFS (i.e. using standard I/O system calls). + * This support is only needed on clients that wish to mount the file system. + * + */ + +/* + * Declarations and macros for the ORANGEFS Linux kernel support. + */ + +#ifndef __ORANGEFSKERNEL_H +#define __ORANGEFSKERNEL_H + +#include <linux/kernel.h> +#include <linux/moduleparam.h> +#include <linux/statfs.h> +#include <linux/backing-dev.h> +#include <linux/device.h> +#include <linux/mpage.h> +#include <linux/namei.h> +#include <linux/errno.h> +#include <linux/init.h> +#include <linux/module.h> +#include <linux/slab.h> +#include <linux/types.h> +#include <linux/fs.h> +#include <linux/vmalloc.h> + +#include <linux/aio.h> +#include <linux/posix_acl.h> +#include <linux/posix_acl_xattr.h> +#include <linux/compat.h> +#include <linux/mount.h> +#include <linux/uaccess.h> +#include <linux/atomic.h> +#include <linux/uio.h> +#include <linux/sched.h> +#include <linux/mm.h> +#include <linux/wait.h> +#include <linux/dcache.h> +#include <linux/pagemap.h> +#include <linux/poll.h> +#include <linux/rwsem.h> +#include <linux/xattr.h> +#include <linux/exportfs.h> + +#include <asm/unaligned.h> + +#include "orangefs-dev-proto.h" + +#ifdef ORANGEFS_KERNEL_DEBUG +#define ORANGEFS_DEFAULT_OP_TIMEOUT_SECS 10 +#else +#define ORANGEFS_DEFAULT_OP_TIMEOUT_SECS 20 +#endif + +#define ORANGEFS_BUFMAP_WAIT_TIMEOUT_SECS 30 + +#define ORANGEFS_DEFAULT_SLOT_TIMEOUT_SECS 900 /* 15 minutes */ + +#define ORANGEFS_REQDEVICE_NAME "pvfs2-req" + +#define ORANGEFS_DEVREQ_MAGIC 0x20030529 +#define ORANGEFS_LINK_MAX 0x000000FF +#define ORANGEFS_PURGE_RETRY_COUNT 0x00000005 +#define ORANGEFS_MAX_NUM_OPTIONS 0x00000004 +#define ORANGEFS_MAX_MOUNT_OPT_LEN 0x00000080 +#define ORANGEFS_MAX_FSKEY_LEN 64 + +#define MAX_DEV_REQ_UPSIZE (2 * sizeof(__s32) + \ +sizeof(__u64) + sizeof(struct orangefs_upcall_s)) +#define MAX_DEV_REQ_DOWNSIZE (2 * sizeof(__s32) + \ +sizeof(__u64) + sizeof(struct orangefs_downcall_s)) + +/* + * valid orangefs kernel operation states + * + * unknown - op was just initialized + * waiting - op is on request_list (upward bound) + * inprogr - op is in progress (waiting for downcall) + * serviced - op has matching downcall; ok + * purged - op has to start a timer since client-core + * exited uncleanly before servicing op + * given up - submitter has given up waiting for it + */ +enum orangefs_vfs_op_states { + OP_VFS_STATE_UNKNOWN = 0, + OP_VFS_STATE_WAITING = 1, + OP_VFS_STATE_INPROGR = 2, + OP_VFS_STATE_SERVICED = 4, + OP_VFS_STATE_PURGED = 8, + OP_VFS_STATE_GIVEN_UP = 16, +}; + +/* + * An array of client_debug_mask will be built to hold debug keyword/mask + * values fetched from userspace. + */ +struct client_debug_mask { + char *keyword; + __u64 mask1; + __u64 mask2; +}; + +/* + * orangefs kernel memory related flags + */ + +#if ((defined ORANGEFS_KERNEL_DEBUG) && (defined CONFIG_DEBUG_SLAB)) +#define ORANGEFS_CACHE_CREATE_FLAGS SLAB_RED_ZONE +#else +#define ORANGEFS_CACHE_CREATE_FLAGS 0 +#endif /* ((defined ORANGEFS_KERNEL_DEBUG) && (defined CONFIG_DEBUG_SLAB)) */ + +/* orangefs xattr and acl related defines */ +#define ORANGEFS_XATTR_INDEX_POSIX_ACL_ACCESS 1 +#define ORANGEFS_XATTR_INDEX_POSIX_ACL_DEFAULT 2 +#define ORANGEFS_XATTR_INDEX_TRUSTED 3 +#define ORANGEFS_XATTR_INDEX_DEFAULT 4 + +#define ORANGEFS_XATTR_NAME_ACL_ACCESS XATTR_NAME_POSIX_ACL_ACCESS +#define ORANGEFS_XATTR_NAME_ACL_DEFAULT XATTR_NAME_POSIX_ACL_DEFAULT +#define ORANGEFS_XATTR_NAME_TRUSTED_PREFIX "trusted." +#define ORANGEFS_XATTR_NAME_DEFAULT_PREFIX "" + +/* these functions are defined in orangefs-utils.c */ +int orangefs_prepare_cdm_array(char *debug_array_string); +int orangefs_prepare_debugfs_help_string(int); + +/* defined in orangefs-debugfs.c */ +int orangefs_client_debug_init(void); + +void debug_string_to_mask(char *, void *, int); +void do_c_mask(int, char *, struct client_debug_mask **); +void do_k_mask(int, char *, __u64 **); + +void debug_mask_to_string(void *, int); +void do_k_string(void *, int); +void do_c_string(void *, int); +int check_amalgam_keyword(void *, int); +int keyword_is_amalgam(char *); + +/*these variables are defined in orangefs-mod.c */ +extern char kernel_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +extern char client_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +extern char client_debug_array_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +extern unsigned int kernel_mask_set_mod_init; + +extern int orangefs_init_acl(struct inode *inode, struct inode *dir); +extern const struct xattr_handler *orangefs_xattr_handlers[]; + +extern struct posix_acl *orangefs_get_acl(struct inode *inode, int type); +extern int orangefs_set_acl(struct inode *inode, struct posix_acl *acl, int type); + +/* + * Redefine xtvec structure so that we could move helper functions out of + * the define + */ +struct xtvec { + __kernel_off_t xtv_off; /* must be off_t */ + __kernel_size_t xtv_len; /* must be size_t */ +}; + +/* + * orangefs data structures + */ +struct orangefs_kernel_op_s { + enum orangefs_vfs_op_states op_state; + __u64 tag; + + /* + * Set uses_shared_memory to non zero if this operation uses + * shared memory. If true, then a retry on the op must also + * get a new shared memory buffer and re-populate it. + * Cancels don't care - it only matters for service_operation() + * retry logics and cancels don't go through it anymore. It + * safely stays non-zero when we use it as slot_to_free. + */ + union { + int uses_shared_memory; + int slot_to_free; + }; + + struct orangefs_upcall_s upcall; + struct orangefs_downcall_s downcall; + + struct completion waitq; + spinlock_t lock; + + int attempts; + + struct list_head list; +}; + +#define set_op_state_waiting(op) ((op)->op_state = OP_VFS_STATE_WAITING) +#define set_op_state_inprogress(op) ((op)->op_state = OP_VFS_STATE_INPROGR) +#define set_op_state_given_up(op) ((op)->op_state = OP_VFS_STATE_GIVEN_UP) +static inline void set_op_state_serviced(struct orangefs_kernel_op_s *op) +{ + op->op_state = OP_VFS_STATE_SERVICED; + complete(&op->waitq); +} + +#define op_state_waiting(op) ((op)->op_state & OP_VFS_STATE_WAITING) +#define op_state_in_progress(op) ((op)->op_state & OP_VFS_STATE_INPROGR) +#define op_state_serviced(op) ((op)->op_state & OP_VFS_STATE_SERVICED) +#define op_state_purged(op) ((op)->op_state & OP_VFS_STATE_PURGED) +#define op_state_given_up(op) ((op)->op_state & OP_VFS_STATE_GIVEN_UP) +#define op_is_cancel(op) ((op)->upcall.type == ORANGEFS_VFS_OP_CANCEL) + +void op_release(struct orangefs_kernel_op_s *op); + +extern void orangefs_bufmap_put(int); +static inline void put_cancel(struct orangefs_kernel_op_s *op) +{ + orangefs_bufmap_put(op->slot_to_free); + op_release(op); +} + +static inline void set_op_state_purged(struct orangefs_kernel_op_s *op) +{ + spin_lock(&op->lock); + if (unlikely(op_is_cancel(op))) { + list_del_init(&op->list); + spin_unlock(&op->lock); + put_cancel(op); + } else { + op->op_state |= OP_VFS_STATE_PURGED; + complete(&op->waitq); + spin_unlock(&op->lock); + } +} + +/* per inode private orangefs info */ +struct orangefs_inode_s { + struct orangefs_object_kref refn; + char link_target[ORANGEFS_NAME_MAX]; + __s64 blksize; + /* + * Reading/Writing Extended attributes need to acquire the appropriate + * reader/writer semaphore on the orangefs_inode_s structure. + */ + struct rw_semaphore xattr_sem; + + struct inode vfs_inode; + sector_t last_failed_block_index_read; + + /* + * State of in-memory attributes not yet flushed to disk associated + * with this object + */ + unsigned long pinode_flags; +}; + +#define P_ATIME_FLAG 0 +#define P_MTIME_FLAG 1 +#define P_CTIME_FLAG 2 +#define P_MODE_FLAG 3 + +#define ClearAtimeFlag(pinode) clear_bit(P_ATIME_FLAG, &(pinode)->pinode_flags) +#define SetAtimeFlag(pinode) set_bit(P_ATIME_FLAG, &(pinode)->pinode_flags) +#define AtimeFlag(pinode) test_bit(P_ATIME_FLAG, &(pinode)->pinode_flags) + +#define ClearMtimeFlag(pinode) clear_bit(P_MTIME_FLAG, &(pinode)->pinode_flags) +#define SetMtimeFlag(pinode) set_bit(P_MTIME_FLAG, &(pinode)->pinode_flags) +#define MtimeFlag(pinode) test_bit(P_MTIME_FLAG, &(pinode)->pinode_flags) + +#define ClearCtimeFlag(pinode) clear_bit(P_CTIME_FLAG, &(pinode)->pinode_flags) +#define SetCtimeFlag(pinode) set_bit(P_CTIME_FLAG, &(pinode)->pinode_flags) +#define CtimeFlag(pinode) test_bit(P_CTIME_FLAG, &(pinode)->pinode_flags) + +#define ClearModeFlag(pinode) clear_bit(P_MODE_FLAG, &(pinode)->pinode_flags) +#define SetModeFlag(pinode) set_bit(P_MODE_FLAG, &(pinode)->pinode_flags) +#define ModeFlag(pinode) test_bit(P_MODE_FLAG, &(pinode)->pinode_flags) + +/* per superblock private orangefs info */ +struct orangefs_sb_info_s { + struct orangefs_khandle root_khandle; + __s32 fs_id; + int id; + int flags; +#define ORANGEFS_OPT_INTR 0x01 +#define ORANGEFS_OPT_LOCAL_LOCK 0x02 + char devname[ORANGEFS_MAX_SERVER_ADDR_LEN]; + struct super_block *sb; + int mount_pending; + struct list_head list; +}; + +/* + * structure that holds the state of any async I/O operation issued + * through the VFS. Needed especially to handle cancellation requests + * or even completion notification so that the VFS client-side daemon + * can free up its vfs_request slots. + */ +struct orangefs_kiocb_s { + /* the pointer to the task that initiated the AIO */ + struct task_struct *tsk; + + /* pointer to the kiocb that kicked this operation */ + struct kiocb *kiocb; + + /* buffer index that was used for the I/O */ + struct orangefs_bufmap *bufmap; + int buffer_index; + + /* orangefs kernel operation type */ + struct orangefs_kernel_op_s *op; + + /* The user space buffers from/to which I/O is being staged */ + struct iovec *iov; + + /* number of elements in the iovector */ + unsigned long nr_segs; + + /* set to indicate the type of the operation */ + int rw; + + /* file offset */ + loff_t offset; + + /* and the count in bytes */ + size_t bytes_to_be_copied; + + ssize_t bytes_copied; + int needs_cleanup; +}; + +struct orangefs_stats { + unsigned long cache_hits; + unsigned long cache_misses; + unsigned long reads; + unsigned long writes; +}; + +extern struct orangefs_stats g_orangefs_stats; + +/* + * NOTE: See Documentation/filesystems/porting for information + * on implementing FOO_I and properly accessing fs private data + */ +static inline struct orangefs_inode_s *ORANGEFS_I(struct inode *inode) +{ + return container_of(inode, struct orangefs_inode_s, vfs_inode); +} + +static inline struct orangefs_sb_info_s *ORANGEFS_SB(struct super_block *sb) +{ + return (struct orangefs_sb_info_s *) sb->s_fs_info; +} + +/* ino_t descends from "unsigned long", 8 bytes, 64 bits. */ +static inline ino_t orangefs_khandle_to_ino(struct orangefs_khandle *khandle) +{ + union { + unsigned char u[8]; + __u64 ino; + } ihandle; + + ihandle.u[0] = khandle->u[0] ^ khandle->u[4]; + ihandle.u[1] = khandle->u[1] ^ khandle->u[5]; + ihandle.u[2] = khandle->u[2] ^ khandle->u[6]; + ihandle.u[3] = khandle->u[3] ^ khandle->u[7]; + ihandle.u[4] = khandle->u[12] ^ khandle->u[8]; + ihandle.u[5] = khandle->u[13] ^ khandle->u[9]; + ihandle.u[6] = khandle->u[14] ^ khandle->u[10]; + ihandle.u[7] = khandle->u[15] ^ khandle->u[11]; + + return ihandle.ino; +} + +static inline struct orangefs_khandle *get_khandle_from_ino(struct inode *inode) +{ + return &(ORANGEFS_I(inode)->refn.khandle); +} + +static inline __s32 get_fsid_from_ino(struct inode *inode) +{ + return ORANGEFS_I(inode)->refn.fs_id; +} + +static inline ino_t get_ino_from_khandle(struct inode *inode) +{ + struct orangefs_khandle *khandle; + ino_t ino; + + khandle = get_khandle_from_ino(inode); + ino = orangefs_khandle_to_ino(khandle); + return ino; +} + +static inline ino_t get_parent_ino_from_dentry(struct dentry *dentry) +{ + return get_ino_from_khandle(dentry->d_parent->d_inode); +} + +static inline int is_root_handle(struct inode *inode) +{ + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s: root handle: %pU, this handle: %pU:\n", + __func__, + &ORANGEFS_SB(inode->i_sb)->root_khandle, + get_khandle_from_ino(inode)); + + if (ORANGEFS_khandle_cmp(&(ORANGEFS_SB(inode->i_sb)->root_khandle), + get_khandle_from_ino(inode))) + return 0; + else + return 1; +} + +static inline int match_handle(struct orangefs_khandle resp_handle, + struct inode *inode) +{ + gossip_debug(GOSSIP_DCACHE_DEBUG, + "%s: one handle: %pU, another handle:%pU:\n", + __func__, + &resp_handle, + get_khandle_from_ino(inode)); + + if (ORANGEFS_khandle_cmp(&resp_handle, get_khandle_from_ino(inode))) + return 0; + else + return 1; +} + +/* + * defined in orangefs-cache.c + */ +int op_cache_initialize(void); +int op_cache_finalize(void); +struct orangefs_kernel_op_s *op_alloc(__s32 type); +void orangefs_new_tag(struct orangefs_kernel_op_s *op); +char *get_opname_string(struct orangefs_kernel_op_s *new_op); + +int orangefs_inode_cache_initialize(void); +int orangefs_inode_cache_finalize(void); + +/* + * defined in orangefs-mod.c + */ +void purge_inprogress_ops(void); + +/* + * defined in waitqueue.c + */ +void purge_waiting_ops(void); + +/* + * defined in super.c + */ +struct dentry *orangefs_mount(struct file_system_type *fst, + int flags, + const char *devname, + void *data); + +void orangefs_kill_sb(struct super_block *sb); +int orangefs_remount(struct orangefs_sb_info_s *); + +int fsid_key_table_initialize(void); +void fsid_key_table_finalize(void); + +/* + * defined in inode.c + */ +__u32 convert_to_orangefs_mask(unsigned long lite_mask); +struct inode *orangefs_new_inode(struct super_block *sb, + struct inode *dir, + int mode, + dev_t dev, + struct orangefs_object_kref *ref); + +int orangefs_setattr(struct dentry *dentry, struct iattr *iattr); + +int orangefs_getattr(struct vfsmount *mnt, + struct dentry *dentry, + struct kstat *kstat); + +int orangefs_permission(struct inode *inode, int mask); + +/* + * defined in xattr.c + */ +int orangefs_setxattr(struct dentry *dentry, + const char *name, + const void *value, + size_t size, + int flags); + +ssize_t orangefs_getxattr(struct dentry *dentry, + const char *name, + void *buffer, + size_t size); + +ssize_t orangefs_listxattr(struct dentry *dentry, char *buffer, size_t size); + +/* + * defined in namei.c + */ +struct inode *orangefs_iget(struct super_block *sb, + struct orangefs_object_kref *ref); + +ssize_t orangefs_inode_read(struct inode *inode, + struct iov_iter *iter, + loff_t *offset, + loff_t readahead_size); + +/* + * defined in devorangefs-req.c + */ +int orangefs_dev_init(void); +void orangefs_dev_cleanup(void); +int is_daemon_in_service(void); +bool __is_daemon_in_service(void); + +/* + * defined in orangefs-utils.c + */ +__s32 fsid_of_op(struct orangefs_kernel_op_s *op); + +int orangefs_flush_inode(struct inode *inode); + +ssize_t orangefs_inode_getxattr(struct inode *inode, + const char *prefix, + const char *name, + void *buffer, + size_t size); + +int orangefs_inode_setxattr(struct inode *inode, + const char *prefix, + const char *name, + const void *value, + size_t size, + int flags); + +int orangefs_inode_getattr(struct inode *inode, int new, int size); + +int orangefs_inode_check_changed(struct inode *inode); + +int orangefs_inode_setattr(struct inode *inode, struct iattr *iattr); + +void orangefs_make_bad_inode(struct inode *inode); + +int orangefs_unmount_sb(struct super_block *sb); + +bool orangefs_cancel_op_in_progress(struct orangefs_kernel_op_s *op); + +int orangefs_normalize_to_errno(__s32 error_code); + +extern struct mutex devreq_mutex; +extern struct mutex request_mutex; +extern int debug; +extern int op_timeout_secs; +extern int slot_timeout_secs; +extern struct list_head orangefs_superblocks; +extern spinlock_t orangefs_superblocks_lock; +extern struct list_head orangefs_request_list; +extern spinlock_t orangefs_request_list_lock; +extern wait_queue_head_t orangefs_request_list_waitq; +extern struct list_head *htable_ops_in_progress; +extern spinlock_t htable_ops_in_progress_lock; +extern int hash_table_size; + +extern const struct address_space_operations orangefs_address_operations; +extern struct backing_dev_info orangefs_backing_dev_info; +extern struct inode_operations orangefs_file_inode_operations; +extern const struct file_operations orangefs_file_operations; +extern struct inode_operations orangefs_symlink_inode_operations; +extern struct inode_operations orangefs_dir_inode_operations; +extern const struct file_operations orangefs_dir_operations; +extern const struct dentry_operations orangefs_dentry_operations; +extern const struct file_operations orangefs_devreq_file_operations; + +extern wait_queue_head_t orangefs_bufmap_init_waitq; + +/* + * misc convenience macros + */ + +#define ORANGEFS_OP_INTERRUPTIBLE 1 /* service_operation() is interruptible */ +#define ORANGEFS_OP_PRIORITY 2 /* service_operation() is high priority */ +#define ORANGEFS_OP_CANCELLATION 4 /* this is a cancellation */ +#define ORANGEFS_OP_NO_MUTEX 8 /* don't acquire request_mutex */ +#define ORANGEFS_OP_ASYNC 16 /* Queue it, but don't wait */ + +int service_operation(struct orangefs_kernel_op_s *op, + const char *op_name, + int flags); + +#define get_interruptible_flag(inode) \ + ((ORANGEFS_SB(inode->i_sb)->flags & ORANGEFS_OPT_INTR) ? \ + ORANGEFS_OP_INTERRUPTIBLE : 0) + +#define fill_default_sys_attrs(sys_attr, type, mode) \ +do { \ + sys_attr.owner = from_kuid(current_user_ns(), current_fsuid()); \ + sys_attr.group = from_kgid(current_user_ns(), current_fsgid()); \ + sys_attr.perms = ORANGEFS_util_translate_mode(mode); \ + sys_attr.mtime = 0; \ + sys_attr.atime = 0; \ + sys_attr.ctime = 0; \ + sys_attr.mask = ORANGEFS_ATTR_SYS_ALL_SETABLE; \ +} while (0) + +static inline void orangefs_i_size_write(struct inode *inode, loff_t i_size) +{ +#if BITS_PER_LONG == 32 && defined(CONFIG_SMP) + mutex_lock(&inode->i_mutex); +#endif + i_size_write(inode, i_size); +#if BITS_PER_LONG == 32 && defined(CONFIG_SMP) + mutex_unlock(&inode->i_mutex); +#endif +} + +#endif /* __ORANGEFSKERNEL_H */ diff --git a/fs/orangefs/orangefs-mod.c b/fs/orangefs/orangefs-mod.c new file mode 100644 index 000000000000..6f072a8c0de1 --- /dev/null +++ b/fs/orangefs/orangefs-mod.c @@ -0,0 +1,293 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * Changes by Acxiom Corporation to add proc file handler for pvfs2 client + * parameters, Copyright Acxiom Corporation, 2005. + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-debugfs.h" +#include "orangefs-sysfs.h" + +/* ORANGEFS_VERSION is a ./configure define */ +#ifndef ORANGEFS_VERSION +#define ORANGEFS_VERSION "upstream" +#endif + +/* + * global variables declared here + */ + +/* array of client debug keyword/mask values */ +struct client_debug_mask *cdm_array; +int cdm_element_count; + +char kernel_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN] = "none"; +char client_debug_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; +char client_debug_array_string[ORANGEFS_MAX_DEBUG_STRING_LEN]; + +char *debug_help_string; +int help_string_initialized; +struct dentry *help_file_dentry; +struct dentry *client_debug_dentry; +struct dentry *debug_dir; +int client_verbose_index; +int client_all_index; +struct orangefs_stats g_orangefs_stats; + +/* the size of the hash tables for ops in progress */ +int hash_table_size = 509; + +static ulong module_parm_debug_mask; +__u64 gossip_debug_mask; +struct client_debug_mask client_debug_mask = { NULL, 0, 0 }; +unsigned int kernel_mask_set_mod_init; /* implicitly false */ +int op_timeout_secs = ORANGEFS_DEFAULT_OP_TIMEOUT_SECS; +int slot_timeout_secs = ORANGEFS_DEFAULT_SLOT_TIMEOUT_SECS; + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("ORANGEFS Development Team"); +MODULE_DESCRIPTION("The Linux Kernel VFS interface to ORANGEFS"); +MODULE_PARM_DESC(module_parm_debug_mask, "debugging level (see orangefs-debug.h for values)"); +MODULE_PARM_DESC(op_timeout_secs, "Operation timeout in seconds"); +MODULE_PARM_DESC(slot_timeout_secs, "Slot timeout in seconds"); +MODULE_PARM_DESC(hash_table_size, + "size of hash table for operations in progress"); + +static struct file_system_type orangefs_fs_type = { + .name = "pvfs2", + .mount = orangefs_mount, + .kill_sb = orangefs_kill_sb, + .owner = THIS_MODULE, +}; + +module_param(hash_table_size, int, 0); +module_param(module_parm_debug_mask, ulong, 0644); +module_param(op_timeout_secs, int, 0); +module_param(slot_timeout_secs, int, 0); + +/* synchronizes the request device file */ +DEFINE_MUTEX(devreq_mutex); + +/* + * Blocks non-priority requests from being queued for servicing. This + * could be used for protecting the request list data structure, but + * for now it's only being used to stall the op addition to the request + * list + */ +DEFINE_MUTEX(request_mutex); + +/* hash table for storing operations waiting for matching downcall */ +struct list_head *htable_ops_in_progress; +DEFINE_SPINLOCK(htable_ops_in_progress_lock); + +/* list for queueing upcall operations */ +LIST_HEAD(orangefs_request_list); + +/* used to protect the above orangefs_request_list */ +DEFINE_SPINLOCK(orangefs_request_list_lock); + +/* used for incoming request notification */ +DECLARE_WAIT_QUEUE_HEAD(orangefs_request_list_waitq); + +static int __init orangefs_init(void) +{ + int ret = -1; + __u32 i = 0; + + /* convert input debug mask to a 64-bit unsigned integer */ + gossip_debug_mask = (unsigned long long) module_parm_debug_mask; + + /* + * set the kernel's gossip debug string; invalid mask values will + * be ignored. + */ + debug_mask_to_string(&gossip_debug_mask, 0); + + /* remove any invalid values from the mask */ + debug_string_to_mask(kernel_debug_string, &gossip_debug_mask, 0); + + /* + * if the mask has a non-zero value, then indicate that the mask + * was set when the kernel module was loaded. The orangefs dev ioctl + * command will look at this boolean to determine if the kernel's + * debug mask should be overwritten when the client-core is started. + */ + if (gossip_debug_mask != 0) + kernel_mask_set_mod_init = true; + + pr_info("%s: called with debug mask: :%s: :%llx:\n", + __func__, + kernel_debug_string, + (unsigned long long)gossip_debug_mask); + + ret = bdi_init(&orangefs_backing_dev_info); + + if (ret) + return ret; + + if (op_timeout_secs < 0) + op_timeout_secs = 0; + + if (slot_timeout_secs < 0) + slot_timeout_secs = 0; + + /* initialize global book keeping data structures */ + ret = op_cache_initialize(); + if (ret < 0) + goto err; + + ret = orangefs_inode_cache_initialize(); + if (ret < 0) + goto cleanup_op; + + htable_ops_in_progress = + kcalloc(hash_table_size, sizeof(struct list_head), GFP_KERNEL); + if (!htable_ops_in_progress) { + gossip_err("Failed to initialize op hashtable"); + ret = -ENOMEM; + goto cleanup_inode; + } + + /* initialize a doubly linked at each hash table index */ + for (i = 0; i < hash_table_size; i++) + INIT_LIST_HEAD(&htable_ops_in_progress[i]); + + ret = fsid_key_table_initialize(); + if (ret < 0) + goto cleanup_progress_table; + + /* + * Build the contents of /sys/kernel/debug/orangefs/debug-help + * from the keywords in the kernel keyword/mask array. + * + * The keywords in the client keyword/mask array are + * unknown at boot time. + * + * orangefs_prepare_debugfs_help_string will be used again + * later to rebuild the debug-help file after the client starts + * and passes along the needed info. The argument signifies + * which time orangefs_prepare_debugfs_help_string is being + * called. + */ + ret = orangefs_prepare_debugfs_help_string(1); + if (ret) + goto cleanup_key_table; + + ret = orangefs_debugfs_init(); + if (ret) + goto debugfs_init_failed; + + ret = orangefs_kernel_debug_init(); + if (ret) + goto kernel_debug_init_failed; + + ret = orangefs_sysfs_init(); + if (ret) + goto sysfs_init_failed; + + /* Initialize the orangefsdev subsystem. */ + ret = orangefs_dev_init(); + if (ret < 0) { + gossip_err("%s: could not initialize device subsystem %d!\n", + __func__, + ret); + goto cleanup_device; + } + + ret = register_filesystem(&orangefs_fs_type); + if (ret == 0) { + pr_info("orangefs: module version %s loaded\n", ORANGEFS_VERSION); + ret = 0; + goto out; + } + + orangefs_sysfs_exit(); + +cleanup_device: + orangefs_dev_cleanup(); + +sysfs_init_failed: + +kernel_debug_init_failed: + +debugfs_init_failed: + orangefs_debugfs_cleanup(); + +cleanup_key_table: + fsid_key_table_finalize(); + +cleanup_progress_table: + kfree(htable_ops_in_progress); + +cleanup_inode: + orangefs_inode_cache_finalize(); + +cleanup_op: + op_cache_finalize(); + +err: + bdi_destroy(&orangefs_backing_dev_info); + +out: + return ret; +} + +static void __exit orangefs_exit(void) +{ + int i = 0; + gossip_debug(GOSSIP_INIT_DEBUG, "orangefs: orangefs_exit called\n"); + + unregister_filesystem(&orangefs_fs_type); + orangefs_debugfs_cleanup(); + orangefs_sysfs_exit(); + fsid_key_table_finalize(); + orangefs_dev_cleanup(); + BUG_ON(!list_empty(&orangefs_request_list)); + for (i = 0; i < hash_table_size; i++) + BUG_ON(!list_empty(&htable_ops_in_progress[i])); + + orangefs_inode_cache_finalize(); + op_cache_finalize(); + + kfree(htable_ops_in_progress); + + bdi_destroy(&orangefs_backing_dev_info); + + pr_info("orangefs: module version %s unloaded\n", ORANGEFS_VERSION); +} + +/* + * What we do in this function is to walk the list of operations + * that are in progress in the hash table and mark them as purged as well. + */ +void purge_inprogress_ops(void) +{ + int i; + + for (i = 0; i < hash_table_size; i++) { + struct orangefs_kernel_op_s *op; + struct orangefs_kernel_op_s *next; + + spin_lock(&htable_ops_in_progress_lock); + list_for_each_entry_safe(op, + next, + &htable_ops_in_progress[i], + list) { + set_op_state_purged(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + } + spin_unlock(&htable_ops_in_progress_lock); + } +} + +module_init(orangefs_init); +module_exit(orangefs_exit); diff --git a/fs/orangefs/orangefs-sysfs.c b/fs/orangefs/orangefs-sysfs.c new file mode 100644 index 000000000000..5c03113e3ad2 --- /dev/null +++ b/fs/orangefs/orangefs-sysfs.c @@ -0,0 +1,1772 @@ +/* + * Documentation/ABI/stable/orangefs-sysfs: + * + * What: /sys/fs/orangefs/perf_counter_reset + * Date: June 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * echo a 0 or a 1 into perf_counter_reset to + * reset all the counters in + * /sys/fs/orangefs/perf_counters + * except ones with PINT_PERF_PRESERVE set. + * + * + * What: /sys/fs/orangefs/perf_counters/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Counters and settings for various caches. + * Read only. + * + * + * What: /sys/fs/orangefs/perf_time_interval_secs + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Length of perf counter intervals in + * seconds. + * + * + * What: /sys/fs/orangefs/perf_history_size + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * The perf_counters cache statistics have N, or + * perf_history_size, samples. The default is + * one. + * + * Every perf_time_interval_secs the (first) + * samples are reset. + * + * If N is greater than one, the "current" set + * of samples is reset, and the samples from the + * other N-1 intervals remain available. + * + * + * What: /sys/fs/orangefs/op_timeout_secs + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Service operation timeout in seconds. + * + * + * What: /sys/fs/orangefs/slot_timeout_secs + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * "Slot" timeout in seconds. A "slot" + * is an indexed buffer in the shared + * memory segment used for communication + * between the kernel module and userspace. + * Slots are requested and waited for, + * the wait times out after slot_timeout_secs. + * + * + * What: /sys/fs/orangefs/acache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Attribute cache configurable settings. + * + * + * What: /sys/fs/orangefs/ncache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Name cache configurable settings. + * + * + * What: /sys/fs/orangefs/capcache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Capability cache configurable settings. + * + * + * What: /sys/fs/orangefs/ccache/... + * Date: Jun 2015 + * Contact: Mike Marshall <hubcap@omnibond.com> + * Description: + * Credential cache configurable settings. + * + */ + +#include <linux/fs.h> +#include <linux/kobject.h> +#include <linux/string.h> +#include <linux/sysfs.h> +#include <linux/module.h> +#include <linux/init.h> + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-sysfs.h" + +#define ORANGEFS_KOBJ_ID "orangefs" +#define ACACHE_KOBJ_ID "acache" +#define CAPCACHE_KOBJ_ID "capcache" +#define CCACHE_KOBJ_ID "ccache" +#define NCACHE_KOBJ_ID "ncache" +#define PC_KOBJ_ID "pc" +#define STATS_KOBJ_ID "stats" + +struct orangefs_obj { + struct kobject kobj; + int op_timeout_secs; + int perf_counter_reset; + int perf_history_size; + int perf_time_interval_secs; + int slot_timeout_secs; +}; + +struct acache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_msecs; +}; + +struct capcache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_secs; +}; + +struct ccache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_secs; +}; + +struct ncache_orangefs_obj { + struct kobject kobj; + int hard_limit; + int reclaim_percentage; + int soft_limit; + int timeout_msecs; +}; + +struct pc_orangefs_obj { + struct kobject kobj; + char *acache; + char *capcache; + char *ncache; +}; + +struct stats_orangefs_obj { + struct kobject kobj; + int reads; + int writes; +}; + +struct orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct acache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct capcache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct capcache_orangefs_obj *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct capcache_orangefs_obj *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct ccache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct ccache_orangefs_obj *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct ccache_orangefs_obj *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct ncache_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct pc_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct pc_orangefs_obj *pc_orangefs_obj, + struct pc_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct pc_orangefs_obj *pc_orangefs_obj, + struct pc_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +struct stats_orangefs_attribute { + struct attribute attr; + ssize_t (*show)(struct stats_orangefs_obj *stats_orangefs_obj, + struct stats_orangefs_attribute *attr, + char *buf); + ssize_t (*store)(struct stats_orangefs_obj *stats_orangefs_obj, + struct stats_orangefs_attribute *attr, + const char *buf, + size_t count); +}; + +static ssize_t orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct orangefs_attribute *attribute; + struct orangefs_obj *orangefs_obj; + int rc; + + attribute = container_of(attr, struct orangefs_attribute, attr); + orangefs_obj = container_of(kobj, struct orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct orangefs_attribute *attribute; + struct orangefs_obj *orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "orangefs_attr_store: start\n"); + + attribute = container_of(attr, struct orangefs_attribute, attr); + orangefs_obj = container_of(kobj, struct orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops orangefs_sysfs_ops = { + .show = orangefs_attr_show, + .store = orangefs_attr_store, +}; + +static ssize_t acache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct acache_orangefs_attribute *attribute; + struct acache_orangefs_obj *acache_orangefs_obj; + int rc; + + attribute = container_of(attr, struct acache_orangefs_attribute, attr); + acache_orangefs_obj = + container_of(kobj, struct acache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(acache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t acache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct acache_orangefs_attribute *attribute; + struct acache_orangefs_obj *acache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "acache_orangefs_attr_store: start\n"); + + attribute = container_of(attr, struct acache_orangefs_attribute, attr); + acache_orangefs_obj = + container_of(kobj, struct acache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(acache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops acache_orangefs_sysfs_ops = { + .show = acache_orangefs_attr_show, + .store = acache_orangefs_attr_store, +}; + +static ssize_t capcache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct capcache_orangefs_attribute *attribute; + struct capcache_orangefs_obj *capcache_orangefs_obj; + int rc; + + attribute = + container_of(attr, struct capcache_orangefs_attribute, attr); + capcache_orangefs_obj = + container_of(kobj, struct capcache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(capcache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t capcache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct capcache_orangefs_attribute *attribute; + struct capcache_orangefs_obj *capcache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "capcache_orangefs_attr_store: start\n"); + + attribute = + container_of(attr, struct capcache_orangefs_attribute, attr); + capcache_orangefs_obj = + container_of(kobj, struct capcache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(capcache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops capcache_orangefs_sysfs_ops = { + .show = capcache_orangefs_attr_show, + .store = capcache_orangefs_attr_store, +}; + +static ssize_t ccache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct ccache_orangefs_attribute *attribute; + struct ccache_orangefs_obj *ccache_orangefs_obj; + int rc; + + attribute = + container_of(attr, struct ccache_orangefs_attribute, attr); + ccache_orangefs_obj = + container_of(kobj, struct ccache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(ccache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t ccache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct ccache_orangefs_attribute *attribute; + struct ccache_orangefs_obj *ccache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "ccache_orangefs_attr_store: start\n"); + + attribute = + container_of(attr, struct ccache_orangefs_attribute, attr); + ccache_orangefs_obj = + container_of(kobj, struct ccache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(ccache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops ccache_orangefs_sysfs_ops = { + .show = ccache_orangefs_attr_show, + .store = ccache_orangefs_attr_store, +}; + +static ssize_t ncache_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct ncache_orangefs_attribute *attribute; + struct ncache_orangefs_obj *ncache_orangefs_obj; + int rc; + + attribute = container_of(attr, struct ncache_orangefs_attribute, attr); + ncache_orangefs_obj = + container_of(kobj, struct ncache_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(ncache_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static ssize_t ncache_orangefs_attr_store(struct kobject *kobj, + struct attribute *attr, + const char *buf, + size_t len) +{ + struct ncache_orangefs_attribute *attribute; + struct ncache_orangefs_obj *ncache_orangefs_obj; + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "ncache_orangefs_attr_store: start\n"); + + attribute = container_of(attr, struct ncache_orangefs_attribute, attr); + ncache_orangefs_obj = + container_of(kobj, struct ncache_orangefs_obj, kobj); + + if (!attribute->store) { + rc = -EIO; + goto out; + } + + rc = attribute->store(ncache_orangefs_obj, attribute, buf, len); + +out: + return rc; +} + +static const struct sysfs_ops ncache_orangefs_sysfs_ops = { + .show = ncache_orangefs_attr_show, + .store = ncache_orangefs_attr_store, +}; + +static ssize_t pc_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct pc_orangefs_attribute *attribute; + struct pc_orangefs_obj *pc_orangefs_obj; + int rc; + + attribute = container_of(attr, struct pc_orangefs_attribute, attr); + pc_orangefs_obj = + container_of(kobj, struct pc_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(pc_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static const struct sysfs_ops pc_orangefs_sysfs_ops = { + .show = pc_orangefs_attr_show, +}; + +static ssize_t stats_orangefs_attr_show(struct kobject *kobj, + struct attribute *attr, + char *buf) +{ + struct stats_orangefs_attribute *attribute; + struct stats_orangefs_obj *stats_orangefs_obj; + int rc; + + attribute = container_of(attr, struct stats_orangefs_attribute, attr); + stats_orangefs_obj = + container_of(kobj, struct stats_orangefs_obj, kobj); + + if (!attribute->show) { + rc = -EIO; + goto out; + } + + rc = attribute->show(stats_orangefs_obj, attribute, buf); + +out: + return rc; +} + +static const struct sysfs_ops stats_orangefs_sysfs_ops = { + .show = stats_orangefs_attr_show, +}; + +static void orangefs_release(struct kobject *kobj) +{ + struct orangefs_obj *orangefs_obj; + + orangefs_obj = container_of(kobj, struct orangefs_obj, kobj); + kfree(orangefs_obj); +} + +static void acache_orangefs_release(struct kobject *kobj) +{ + struct acache_orangefs_obj *acache_orangefs_obj; + + acache_orangefs_obj = + container_of(kobj, struct acache_orangefs_obj, kobj); + kfree(acache_orangefs_obj); +} + +static void capcache_orangefs_release(struct kobject *kobj) +{ + struct capcache_orangefs_obj *capcache_orangefs_obj; + + capcache_orangefs_obj = + container_of(kobj, struct capcache_orangefs_obj, kobj); + kfree(capcache_orangefs_obj); +} + +static void ccache_orangefs_release(struct kobject *kobj) +{ + struct ccache_orangefs_obj *ccache_orangefs_obj; + + ccache_orangefs_obj = + container_of(kobj, struct ccache_orangefs_obj, kobj); + kfree(ccache_orangefs_obj); +} + +static void ncache_orangefs_release(struct kobject *kobj) +{ + struct ncache_orangefs_obj *ncache_orangefs_obj; + + ncache_orangefs_obj = + container_of(kobj, struct ncache_orangefs_obj, kobj); + kfree(ncache_orangefs_obj); +} + +static void pc_orangefs_release(struct kobject *kobj) +{ + struct pc_orangefs_obj *pc_orangefs_obj; + + pc_orangefs_obj = + container_of(kobj, struct pc_orangefs_obj, kobj); + kfree(pc_orangefs_obj); +} + +static void stats_orangefs_release(struct kobject *kobj) +{ + struct stats_orangefs_obj *stats_orangefs_obj; + + stats_orangefs_obj = + container_of(kobj, struct stats_orangefs_obj, kobj); + kfree(stats_orangefs_obj); +} + +static ssize_t sysfs_int_show(char *kobj_id, char *buf, void *attr) +{ + int rc = -EIO; + struct orangefs_attribute *orangefs_attr; + struct stats_orangefs_attribute *stats_orangefs_attr; + + gossip_debug(GOSSIP_SYSFS_DEBUG, "sysfs_int_show: id:%s:\n", kobj_id); + + if (!strcmp(kobj_id, ORANGEFS_KOBJ_ID)) { + orangefs_attr = (struct orangefs_attribute *)attr; + + if (!strcmp(orangefs_attr->attr.name, "op_timeout_secs")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%d\n", + op_timeout_secs); + goto out; + } else if (!strcmp(orangefs_attr->attr.name, + "slot_timeout_secs")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%d\n", + slot_timeout_secs); + goto out; + } else { + goto out; + } + + } else if (!strcmp(kobj_id, STATS_KOBJ_ID)) { + stats_orangefs_attr = (struct stats_orangefs_attribute *)attr; + + if (!strcmp(stats_orangefs_attr->attr.name, "reads")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%lu\n", + g_orangefs_stats.reads); + goto out; + } else if (!strcmp(stats_orangefs_attr->attr.name, "writes")) { + rc = scnprintf(buf, + PAGE_SIZE, + "%lu\n", + g_orangefs_stats.writes); + goto out; + } else { + goto out; + } + } + +out: + + return rc; +} + +static ssize_t int_orangefs_show(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + char *buf) +{ + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "int_orangefs_show:start attr->attr.name:%s:\n", + attr->attr.name); + + rc = sysfs_int_show(ORANGEFS_KOBJ_ID, buf, (void *) attr); + + return rc; +} + +static ssize_t int_stats_show(struct stats_orangefs_obj *stats_orangefs_obj, + struct stats_orangefs_attribute *attr, + char *buf) +{ + int rc; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "int_stats_show:start attr->attr.name:%s:\n", + attr->attr.name); + + rc = sysfs_int_show(STATS_KOBJ_ID, buf, (void *) attr); + + return rc; +} + +static ssize_t int_store(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "int_store: start attr->attr.name:%s: buf:%s:\n", + attr->attr.name, buf); + + if (!strcmp(attr->attr.name, "op_timeout_secs")) { + rc = kstrtoint(buf, 0, &op_timeout_secs); + goto out; + } else if (!strcmp(attr->attr.name, "slot_timeout_secs")) { + rc = kstrtoint(buf, 0, &slot_timeout_secs); + goto out; + } else { + goto out; + } + +out: + if (rc) + rc = -EINVAL; + else + rc = count; + + return rc; +} + +/* + * obtain attribute values from userspace with a service operation. + */ +static int sysfs_service_op_show(char *kobj_id, char *buf, void *attr) +{ + struct orangefs_kernel_op_s *new_op = NULL; + int rc = 0; + char *ser_op_type = NULL; + struct orangefs_attribute *orangefs_attr; + struct acache_orangefs_attribute *acache_attr; + struct capcache_orangefs_attribute *capcache_attr; + struct ccache_orangefs_attribute *ccache_attr; + struct ncache_orangefs_attribute *ncache_attr; + struct pc_orangefs_attribute *pc_attr; + __u32 op_alloc_type; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "sysfs_service_op_show: id:%s:\n", + kobj_id); + + if (strcmp(kobj_id, PC_KOBJ_ID)) + op_alloc_type = ORANGEFS_VFS_OP_PARAM; + else + op_alloc_type = ORANGEFS_VFS_OP_PERF_COUNT; + + new_op = op_alloc(op_alloc_type); + if (!new_op) + return -ENOMEM; + + /* Can't do a service_operation if the client is not running... */ + rc = is_daemon_in_service(); + if (rc) { + pr_info("%s: Client not running :%d:\n", + __func__, + is_daemon_in_service()); + goto out; + } + + if (strcmp(kobj_id, PC_KOBJ_ID)) + new_op->upcall.req.param.type = ORANGEFS_PARAM_REQUEST_GET; + + if (!strcmp(kobj_id, ORANGEFS_KOBJ_ID)) { + orangefs_attr = (struct orangefs_attribute *)attr; + + if (!strcmp(orangefs_attr->attr.name, "perf_history_size")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_HISTORY_SIZE; + else if (!strcmp(orangefs_attr->attr.name, + "perf_time_interval_secs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_TIME_INTERVAL_SECS; + else if (!strcmp(orangefs_attr->attr.name, + "perf_counter_reset")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_RESET; + + } else if (!strcmp(kobj_id, ACACHE_KOBJ_ID)) { + acache_attr = (struct acache_orangefs_attribute *)attr; + + if (!strcmp(acache_attr->attr.name, "timeout_msecs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_TIMEOUT_MSECS; + + if (!strcmp(acache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_HARD_LIMIT; + + if (!strcmp(acache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_SOFT_LIMIT; + + if (!strcmp(acache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, CAPCACHE_KOBJ_ID)) { + capcache_attr = (struct capcache_orangefs_attribute *)attr; + + if (!strcmp(capcache_attr->attr.name, "timeout_secs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_TIMEOUT_SECS; + + if (!strcmp(capcache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_HARD_LIMIT; + + if (!strcmp(capcache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_SOFT_LIMIT; + + if (!strcmp(capcache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, CCACHE_KOBJ_ID)) { + ccache_attr = (struct ccache_orangefs_attribute *)attr; + + if (!strcmp(ccache_attr->attr.name, "timeout_secs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_TIMEOUT_SECS; + + if (!strcmp(ccache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_HARD_LIMIT; + + if (!strcmp(ccache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_SOFT_LIMIT; + + if (!strcmp(ccache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, NCACHE_KOBJ_ID)) { + ncache_attr = (struct ncache_orangefs_attribute *)attr; + + if (!strcmp(ncache_attr->attr.name, "timeout_msecs")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_TIMEOUT_MSECS; + + if (!strcmp(ncache_attr->attr.name, "hard_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_HARD_LIMIT; + + if (!strcmp(ncache_attr->attr.name, "soft_limit")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_SOFT_LIMIT; + + if (!strcmp(ncache_attr->attr.name, "reclaim_percentage")) + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_RECLAIM_PERCENTAGE; + + } else if (!strcmp(kobj_id, PC_KOBJ_ID)) { + pc_attr = (struct pc_orangefs_attribute *)attr; + + if (!strcmp(pc_attr->attr.name, ACACHE_KOBJ_ID)) + new_op->upcall.req.perf_count.type = + ORANGEFS_PERF_COUNT_REQUEST_ACACHE; + + if (!strcmp(pc_attr->attr.name, CAPCACHE_KOBJ_ID)) + new_op->upcall.req.perf_count.type = + ORANGEFS_PERF_COUNT_REQUEST_CAPCACHE; + + if (!strcmp(pc_attr->attr.name, NCACHE_KOBJ_ID)) + new_op->upcall.req.perf_count.type = + ORANGEFS_PERF_COUNT_REQUEST_NCACHE; + + } else { + gossip_err("sysfs_service_op_show: unknown kobj_id:%s:\n", + kobj_id); + rc = -EINVAL; + goto out; + } + + + if (strcmp(kobj_id, PC_KOBJ_ID)) + ser_op_type = "orangefs_param"; + else + ser_op_type = "orangefs_perf_count"; + + /* + * The service_operation will return an errno return code on + * error, and zero on success. + */ + rc = service_operation(new_op, ser_op_type, ORANGEFS_OP_INTERRUPTIBLE); + +out: + if (!rc) { + if (strcmp(kobj_id, PC_KOBJ_ID)) { + rc = scnprintf(buf, + PAGE_SIZE, + "%d\n", + (int)new_op->downcall.resp.param.value); + } else { + rc = scnprintf( + buf, + PAGE_SIZE, + "%s", + new_op->downcall.resp.perf_count.buffer); + } + } + + op_release(new_op); + + return rc; + +} + +static ssize_t service_orangefs_show(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(ORANGEFS_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t + service_acache_show(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(ACACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t service_capcache_show(struct capcache_orangefs_obj + *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(CAPCACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t service_ccache_show(struct ccache_orangefs_obj + *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(CCACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t + service_ncache_show(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(NCACHE_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +static ssize_t + service_pc_show(struct pc_orangefs_obj *pc_orangefs_obj, + struct pc_orangefs_attribute *attr, + char *buf) +{ + int rc = 0; + + rc = sysfs_service_op_show(PC_KOBJ_ID, buf, (void *)attr); + + return rc; +} + +/* + * pass attribute values back to userspace with a service operation. + * + * We have to do a memory allocation, an sscanf and a service operation. + * And we have to evaluate what the user entered, to make sure the + * value is within the range supported by the attribute. So, there's + * a lot of return code checking and mapping going on here. + * + * We want to return 1 if we think everything went OK, and + * EINVAL if not. + */ +static int sysfs_service_op_store(char *kobj_id, const char *buf, void *attr) +{ + struct orangefs_kernel_op_s *new_op = NULL; + int val = 0; + int rc = 0; + struct orangefs_attribute *orangefs_attr; + struct acache_orangefs_attribute *acache_attr; + struct capcache_orangefs_attribute *capcache_attr; + struct ccache_orangefs_attribute *ccache_attr; + struct ncache_orangefs_attribute *ncache_attr; + + gossip_debug(GOSSIP_SYSFS_DEBUG, + "sysfs_service_op_store: id:%s:\n", + kobj_id); + + new_op = op_alloc(ORANGEFS_VFS_OP_PARAM); + if (!new_op) + return -EINVAL; /* sic */ + + /* Can't do a service_operation if the client is not running... */ + rc = is_daemon_in_service(); + if (rc) { + pr_info("%s: Client not running :%d:\n", + __func__, + is_daemon_in_service()); + goto out; + } + + /* + * The value we want to send back to userspace is in buf. + */ + rc = kstrtoint(buf, 0, &val); + if (rc) + goto out; + + if (!strcmp(kobj_id, ORANGEFS_KOBJ_ID)) { + orangefs_attr = (struct orangefs_attribute *)attr; + + if (!strcmp(orangefs_attr->attr.name, "perf_history_size")) { + if (val > 0) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_HISTORY_SIZE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(orangefs_attr->attr.name, + "perf_time_interval_secs")) { + if (val > 0) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_TIME_INTERVAL_SECS; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(orangefs_attr->attr.name, + "perf_counter_reset")) { + if ((val == 0) || (val == 1)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_PERF_RESET; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, ACACHE_KOBJ_ID)) { + acache_attr = (struct acache_orangefs_attribute *)attr; + + if (!strcmp(acache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(acache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(acache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(acache_attr->attr.name, "timeout_msecs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_ACACHE_TIMEOUT_MSECS; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, CAPCACHE_KOBJ_ID)) { + capcache_attr = (struct capcache_orangefs_attribute *)attr; + + if (!strcmp(capcache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(capcache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(capcache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(capcache_attr->attr.name, "timeout_secs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_TIMEOUT_SECS; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, CCACHE_KOBJ_ID)) { + ccache_attr = (struct ccache_orangefs_attribute *)attr; + + if (!strcmp(ccache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ccache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ccache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ccache_attr->attr.name, "timeout_secs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_CCACHE_TIMEOUT_SECS; + } else { + rc = 0; + goto out; + } + } + + } else if (!strcmp(kobj_id, NCACHE_KOBJ_ID)) { + ncache_attr = (struct ncache_orangefs_attribute *)attr; + + if (!strcmp(ncache_attr->attr.name, "hard_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_HARD_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ncache_attr->attr.name, "soft_limit")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_SOFT_LIMIT; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ncache_attr->attr.name, + "reclaim_percentage")) { + if ((val > -1) && (val < 101)) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_RECLAIM_PERCENTAGE; + } else { + rc = 0; + goto out; + } + } else if (!strcmp(ncache_attr->attr.name, "timeout_msecs")) { + if (val > -1) { + new_op->upcall.req.param.op = + ORANGEFS_PARAM_REQUEST_OP_NCACHE_TIMEOUT_MSECS; + } else { + rc = 0; + goto out; + } + } + + } else { + gossip_err("sysfs_service_op_store: unknown kobj_id:%s:\n", + kobj_id); + rc = -EINVAL; + goto out; + } + + new_op->upcall.req.param.type = ORANGEFS_PARAM_REQUEST_SET; + + new_op->upcall.req.param.value = val; + + /* + * The service_operation will return a errno return code on + * error, and zero on success. + */ + rc = service_operation(new_op, "orangefs_param", ORANGEFS_OP_INTERRUPTIBLE); + + if (rc < 0) { + gossip_err("sysfs_service_op_store: service op returned:%d:\n", + rc); + rc = 0; + } else { + rc = 1; + } + +out: + op_release(new_op); + + if (rc == -ENOMEM || rc == 0) + rc = -EINVAL; + + return rc; +} + +static ssize_t + service_orangefs_store(struct orangefs_obj *orangefs_obj, + struct orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(ORANGEFS_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t + service_acache_store(struct acache_orangefs_obj *acache_orangefs_obj, + struct acache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(ACACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t + service_capcache_store(struct capcache_orangefs_obj + *capcache_orangefs_obj, + struct capcache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(CAPCACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t service_ccache_store(struct ccache_orangefs_obj + *ccache_orangefs_obj, + struct ccache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(CCACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static ssize_t + service_ncache_store(struct ncache_orangefs_obj *ncache_orangefs_obj, + struct ncache_orangefs_attribute *attr, + const char *buf, + size_t count) +{ + int rc = 0; + + rc = sysfs_service_op_store(NCACHE_KOBJ_ID, buf, (void *) attr); + + /* rc should have an errno value if the service_op went bad. */ + if (rc == 1) + rc = count; + + return rc; +} + +static struct orangefs_attribute op_timeout_secs_attribute = + __ATTR(op_timeout_secs, 0664, int_orangefs_show, int_store); + +static struct orangefs_attribute slot_timeout_secs_attribute = + __ATTR(slot_timeout_secs, 0664, int_orangefs_show, int_store); + +static struct orangefs_attribute perf_counter_reset_attribute = + __ATTR(perf_counter_reset, + 0664, + service_orangefs_show, + service_orangefs_store); + +static struct orangefs_attribute perf_history_size_attribute = + __ATTR(perf_history_size, + 0664, + service_orangefs_show, + service_orangefs_store); + +static struct orangefs_attribute perf_time_interval_secs_attribute = + __ATTR(perf_time_interval_secs, + 0664, + service_orangefs_show, + service_orangefs_store); + +static struct attribute *orangefs_default_attrs[] = { + &op_timeout_secs_attribute.attr, + &slot_timeout_secs_attribute.attr, + &perf_counter_reset_attribute.attr, + &perf_history_size_attribute.attr, + &perf_time_interval_secs_attribute.attr, + NULL, +}; + +static struct kobj_type orangefs_ktype = { + .sysfs_ops = &orangefs_sysfs_ops, + .release = orangefs_release, + .default_attrs = orangefs_default_attrs, +}; + +static struct acache_orangefs_attribute acache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_acache_show, + service_acache_store); + +static struct acache_orangefs_attribute acache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_acache_show, + service_acache_store); + +static struct acache_orangefs_attribute acache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_acache_show, + service_acache_store); + +static struct acache_orangefs_attribute acache_timeout_msecs_attribute = + __ATTR(timeout_msecs, + 0664, + service_acache_show, + service_acache_store); + +static struct attribute *acache_orangefs_default_attrs[] = { + &acache_hard_limit_attribute.attr, + &acache_reclaim_percent_attribute.attr, + &acache_soft_limit_attribute.attr, + &acache_timeout_msecs_attribute.attr, + NULL, +}; + +static struct kobj_type acache_orangefs_ktype = { + .sysfs_ops = &acache_orangefs_sysfs_ops, + .release = acache_orangefs_release, + .default_attrs = acache_orangefs_default_attrs, +}; + +static struct capcache_orangefs_attribute capcache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_capcache_show, + service_capcache_store); + +static struct capcache_orangefs_attribute capcache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_capcache_show, + service_capcache_store); + +static struct capcache_orangefs_attribute capcache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_capcache_show, + service_capcache_store); + +static struct capcache_orangefs_attribute capcache_timeout_secs_attribute = + __ATTR(timeout_secs, + 0664, + service_capcache_show, + service_capcache_store); + +static struct attribute *capcache_orangefs_default_attrs[] = { + &capcache_hard_limit_attribute.attr, + &capcache_reclaim_percent_attribute.attr, + &capcache_soft_limit_attribute.attr, + &capcache_timeout_secs_attribute.attr, + NULL, +}; + +static struct kobj_type capcache_orangefs_ktype = { + .sysfs_ops = &capcache_orangefs_sysfs_ops, + .release = capcache_orangefs_release, + .default_attrs = capcache_orangefs_default_attrs, +}; + +static struct ccache_orangefs_attribute ccache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_ccache_show, + service_ccache_store); + +static struct ccache_orangefs_attribute ccache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_ccache_show, + service_ccache_store); + +static struct ccache_orangefs_attribute ccache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_ccache_show, + service_ccache_store); + +static struct ccache_orangefs_attribute ccache_timeout_secs_attribute = + __ATTR(timeout_secs, + 0664, + service_ccache_show, + service_ccache_store); + +static struct attribute *ccache_orangefs_default_attrs[] = { + &ccache_hard_limit_attribute.attr, + &ccache_reclaim_percent_attribute.attr, + &ccache_soft_limit_attribute.attr, + &ccache_timeout_secs_attribute.attr, + NULL, +}; + +static struct kobj_type ccache_orangefs_ktype = { + .sysfs_ops = &ccache_orangefs_sysfs_ops, + .release = ccache_orangefs_release, + .default_attrs = ccache_orangefs_default_attrs, +}; + +static struct ncache_orangefs_attribute ncache_hard_limit_attribute = + __ATTR(hard_limit, + 0664, + service_ncache_show, + service_ncache_store); + +static struct ncache_orangefs_attribute ncache_reclaim_percent_attribute = + __ATTR(reclaim_percentage, + 0664, + service_ncache_show, + service_ncache_store); + +static struct ncache_orangefs_attribute ncache_soft_limit_attribute = + __ATTR(soft_limit, + 0664, + service_ncache_show, + service_ncache_store); + +static struct ncache_orangefs_attribute ncache_timeout_msecs_attribute = + __ATTR(timeout_msecs, + 0664, + service_ncache_show, + service_ncache_store); + +static struct attribute *ncache_orangefs_default_attrs[] = { + &ncache_hard_limit_attribute.attr, + &ncache_reclaim_percent_attribute.attr, + &ncache_soft_limit_attribute.attr, + &ncache_timeout_msecs_attribute.attr, + NULL, +}; + +static struct kobj_type ncache_orangefs_ktype = { + .sysfs_ops = &ncache_orangefs_sysfs_ops, + .release = ncache_orangefs_release, + .default_attrs = ncache_orangefs_default_attrs, +}; + +static struct pc_orangefs_attribute pc_acache_attribute = + __ATTR(acache, + 0664, + service_pc_show, + NULL); + +static struct pc_orangefs_attribute pc_capcache_attribute = + __ATTR(capcache, + 0664, + service_pc_show, + NULL); + +static struct pc_orangefs_attribute pc_ncache_attribute = + __ATTR(ncache, + 0664, + service_pc_show, + NULL); + +static struct attribute *pc_orangefs_default_attrs[] = { + &pc_acache_attribute.attr, + &pc_capcache_attribute.attr, + &pc_ncache_attribute.attr, + NULL, +}; + +static struct kobj_type pc_orangefs_ktype = { + .sysfs_ops = &pc_orangefs_sysfs_ops, + .release = pc_orangefs_release, + .default_attrs = pc_orangefs_default_attrs, +}; + +static struct stats_orangefs_attribute stats_reads_attribute = + __ATTR(reads, + 0664, + int_stats_show, + NULL); + +static struct stats_orangefs_attribute stats_writes_attribute = + __ATTR(writes, + 0664, + int_stats_show, + NULL); + +static struct attribute *stats_orangefs_default_attrs[] = { + &stats_reads_attribute.attr, + &stats_writes_attribute.attr, + NULL, +}; + +static struct kobj_type stats_orangefs_ktype = { + .sysfs_ops = &stats_orangefs_sysfs_ops, + .release = stats_orangefs_release, + .default_attrs = stats_orangefs_default_attrs, +}; + +static struct orangefs_obj *orangefs_obj; +static struct acache_orangefs_obj *acache_orangefs_obj; +static struct capcache_orangefs_obj *capcache_orangefs_obj; +static struct ccache_orangefs_obj *ccache_orangefs_obj; +static struct ncache_orangefs_obj *ncache_orangefs_obj; +static struct pc_orangefs_obj *pc_orangefs_obj; +static struct stats_orangefs_obj *stats_orangefs_obj; + +int orangefs_sysfs_init(void) +{ + int rc = -EINVAL; + + gossip_debug(GOSSIP_SYSFS_DEBUG, "orangefs_sysfs_init: start\n"); + + /* create /sys/fs/orangefs. */ + orangefs_obj = kzalloc(sizeof(*orangefs_obj), GFP_KERNEL); + if (!orangefs_obj) + goto out; + + rc = kobject_init_and_add(&orangefs_obj->kobj, + &orangefs_ktype, + fs_kobj, + ORANGEFS_KOBJ_ID); + + if (rc) + goto ofs_obj_bail; + + kobject_uevent(&orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/acache. */ + acache_orangefs_obj = kzalloc(sizeof(*acache_orangefs_obj), GFP_KERNEL); + if (!acache_orangefs_obj) { + rc = -EINVAL; + goto ofs_obj_bail; + } + + rc = kobject_init_and_add(&acache_orangefs_obj->kobj, + &acache_orangefs_ktype, + &orangefs_obj->kobj, + ACACHE_KOBJ_ID); + + if (rc) + goto acache_obj_bail; + + kobject_uevent(&acache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/capcache. */ + capcache_orangefs_obj = + kzalloc(sizeof(*capcache_orangefs_obj), GFP_KERNEL); + if (!capcache_orangefs_obj) { + rc = -EINVAL; + goto acache_obj_bail; + } + + rc = kobject_init_and_add(&capcache_orangefs_obj->kobj, + &capcache_orangefs_ktype, + &orangefs_obj->kobj, + CAPCACHE_KOBJ_ID); + if (rc) + goto capcache_obj_bail; + + kobject_uevent(&capcache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/ccache. */ + ccache_orangefs_obj = + kzalloc(sizeof(*ccache_orangefs_obj), GFP_KERNEL); + if (!ccache_orangefs_obj) { + rc = -EINVAL; + goto capcache_obj_bail; + } + + rc = kobject_init_and_add(&ccache_orangefs_obj->kobj, + &ccache_orangefs_ktype, + &orangefs_obj->kobj, + CCACHE_KOBJ_ID); + if (rc) + goto ccache_obj_bail; + + kobject_uevent(&ccache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/ncache. */ + ncache_orangefs_obj = kzalloc(sizeof(*ncache_orangefs_obj), GFP_KERNEL); + if (!ncache_orangefs_obj) { + rc = -EINVAL; + goto ccache_obj_bail; + } + + rc = kobject_init_and_add(&ncache_orangefs_obj->kobj, + &ncache_orangefs_ktype, + &orangefs_obj->kobj, + NCACHE_KOBJ_ID); + + if (rc) + goto ncache_obj_bail; + + kobject_uevent(&ncache_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/perf_counters. */ + pc_orangefs_obj = kzalloc(sizeof(*pc_orangefs_obj), GFP_KERNEL); + if (!pc_orangefs_obj) { + rc = -EINVAL; + goto ncache_obj_bail; + } + + rc = kobject_init_and_add(&pc_orangefs_obj->kobj, + &pc_orangefs_ktype, + &orangefs_obj->kobj, + "perf_counters"); + + if (rc) + goto pc_obj_bail; + + kobject_uevent(&pc_orangefs_obj->kobj, KOBJ_ADD); + + /* create /sys/fs/orangefs/stats. */ + stats_orangefs_obj = kzalloc(sizeof(*stats_orangefs_obj), GFP_KERNEL); + if (!stats_orangefs_obj) { + rc = -EINVAL; + goto pc_obj_bail; + } + + rc = kobject_init_and_add(&stats_orangefs_obj->kobj, + &stats_orangefs_ktype, + &orangefs_obj->kobj, + STATS_KOBJ_ID); + + if (rc) + goto stats_obj_bail; + + kobject_uevent(&stats_orangefs_obj->kobj, KOBJ_ADD); + goto out; + +stats_obj_bail: + kobject_put(&stats_orangefs_obj->kobj); + +pc_obj_bail: + kobject_put(&pc_orangefs_obj->kobj); + +ncache_obj_bail: + kobject_put(&ncache_orangefs_obj->kobj); + +ccache_obj_bail: + kobject_put(&ccache_orangefs_obj->kobj); + +capcache_obj_bail: + kobject_put(&capcache_orangefs_obj->kobj); + +acache_obj_bail: + kobject_put(&acache_orangefs_obj->kobj); + +ofs_obj_bail: + kobject_put(&orangefs_obj->kobj); +out: + return rc; +} + +void orangefs_sysfs_exit(void) +{ + gossip_debug(GOSSIP_SYSFS_DEBUG, "orangefs_sysfs_exit: start\n"); + + kobject_put(&acache_orangefs_obj->kobj); + kobject_put(&capcache_orangefs_obj->kobj); + kobject_put(&ccache_orangefs_obj->kobj); + kobject_put(&ncache_orangefs_obj->kobj); + kobject_put(&pc_orangefs_obj->kobj); + kobject_put(&stats_orangefs_obj->kobj); + + kobject_put(&orangefs_obj->kobj); +} diff --git a/fs/orangefs/orangefs-sysfs.h b/fs/orangefs/orangefs-sysfs.h new file mode 100644 index 000000000000..f0b76382db02 --- /dev/null +++ b/fs/orangefs/orangefs-sysfs.h @@ -0,0 +1,2 @@ +extern int orangefs_sysfs_init(void); +extern void orangefs_sysfs_exit(void); diff --git a/fs/orangefs/orangefs-utils.c b/fs/orangefs/orangefs-utils.c new file mode 100644 index 000000000000..40f5163b56aa --- /dev/null +++ b/fs/orangefs/orangefs-utils.c @@ -0,0 +1,1048 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-dev-proto.h" +#include "orangefs-bufmap.h" + +__s32 fsid_of_op(struct orangefs_kernel_op_s *op) +{ + __s32 fsid = ORANGEFS_FS_ID_NULL; + + if (op) { + switch (op->upcall.type) { + case ORANGEFS_VFS_OP_FILE_IO: + fsid = op->upcall.req.io.refn.fs_id; + break; + case ORANGEFS_VFS_OP_LOOKUP: + fsid = op->upcall.req.lookup.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_CREATE: + fsid = op->upcall.req.create.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_GETATTR: + fsid = op->upcall.req.getattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_REMOVE: + fsid = op->upcall.req.remove.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_MKDIR: + fsid = op->upcall.req.mkdir.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_READDIR: + fsid = op->upcall.req.readdir.refn.fs_id; + break; + case ORANGEFS_VFS_OP_SETATTR: + fsid = op->upcall.req.setattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_SYMLINK: + fsid = op->upcall.req.sym.parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_RENAME: + fsid = op->upcall.req.rename.old_parent_refn.fs_id; + break; + case ORANGEFS_VFS_OP_STATFS: + fsid = op->upcall.req.statfs.fs_id; + break; + case ORANGEFS_VFS_OP_TRUNCATE: + fsid = op->upcall.req.truncate.refn.fs_id; + break; + case ORANGEFS_VFS_OP_MMAP_RA_FLUSH: + fsid = op->upcall.req.ra_cache_flush.refn.fs_id; + break; + case ORANGEFS_VFS_OP_FS_UMOUNT: + fsid = op->upcall.req.fs_umount.fs_id; + break; + case ORANGEFS_VFS_OP_GETXATTR: + fsid = op->upcall.req.getxattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_SETXATTR: + fsid = op->upcall.req.setxattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_LISTXATTR: + fsid = op->upcall.req.listxattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_REMOVEXATTR: + fsid = op->upcall.req.removexattr.refn.fs_id; + break; + case ORANGEFS_VFS_OP_FSYNC: + fsid = op->upcall.req.fsync.refn.fs_id; + break; + default: + break; + } + } + return fsid; +} + +static int orangefs_inode_flags(struct ORANGEFS_sys_attr_s *attrs) +{ + int flags = 0; + if (attrs->flags & ORANGEFS_IMMUTABLE_FL) + flags |= S_IMMUTABLE; + else + flags &= ~S_IMMUTABLE; + if (attrs->flags & ORANGEFS_APPEND_FL) + flags |= S_APPEND; + else + flags &= ~S_APPEND; + if (attrs->flags & ORANGEFS_NOATIME_FL) + flags |= S_NOATIME; + else + flags &= ~S_NOATIME; + return flags; +} + +static int orangefs_inode_perms(struct ORANGEFS_sys_attr_s *attrs) +{ + int perm_mode = 0; + + if (attrs->perms & ORANGEFS_O_EXECUTE) + perm_mode |= S_IXOTH; + if (attrs->perms & ORANGEFS_O_WRITE) + perm_mode |= S_IWOTH; + if (attrs->perms & ORANGEFS_O_READ) + perm_mode |= S_IROTH; + + if (attrs->perms & ORANGEFS_G_EXECUTE) + perm_mode |= S_IXGRP; + if (attrs->perms & ORANGEFS_G_WRITE) + perm_mode |= S_IWGRP; + if (attrs->perms & ORANGEFS_G_READ) + perm_mode |= S_IRGRP; + + if (attrs->perms & ORANGEFS_U_EXECUTE) + perm_mode |= S_IXUSR; + if (attrs->perms & ORANGEFS_U_WRITE) + perm_mode |= S_IWUSR; + if (attrs->perms & ORANGEFS_U_READ) + perm_mode |= S_IRUSR; + + if (attrs->perms & ORANGEFS_G_SGID) + perm_mode |= S_ISGID; + if (attrs->perms & ORANGEFS_U_SUID) + perm_mode |= S_ISUID; + + return perm_mode; +} + +/* + * NOTE: in kernel land, we never use the sys_attr->link_target for + * anything, so don't bother copying it into the sys_attr object here. + */ +static inline int copy_attributes_from_inode(struct inode *inode, + struct ORANGEFS_sys_attr_s *attrs, + struct iattr *iattr) +{ + umode_t tmp_mode; + + if (!iattr || !inode || !attrs) { + gossip_err("NULL iattr (%p), inode (%p), attrs (%p) " + "in copy_attributes_from_inode!\n", + iattr, + inode, + attrs); + return -EINVAL; + } + /* + * We need to be careful to only copy the attributes out of the + * iattr object that we know are valid. + */ + attrs->mask = 0; + if (iattr->ia_valid & ATTR_UID) { + attrs->owner = from_kuid(current_user_ns(), iattr->ia_uid); + attrs->mask |= ORANGEFS_ATTR_SYS_UID; + gossip_debug(GOSSIP_UTILS_DEBUG, "(UID) %d\n", attrs->owner); + } + if (iattr->ia_valid & ATTR_GID) { + attrs->group = from_kgid(current_user_ns(), iattr->ia_gid); + attrs->mask |= ORANGEFS_ATTR_SYS_GID; + gossip_debug(GOSSIP_UTILS_DEBUG, "(GID) %d\n", attrs->group); + } + + if (iattr->ia_valid & ATTR_ATIME) { + attrs->mask |= ORANGEFS_ATTR_SYS_ATIME; + if (iattr->ia_valid & ATTR_ATIME_SET) { + attrs->atime = (time64_t)iattr->ia_atime.tv_sec; + attrs->mask |= ORANGEFS_ATTR_SYS_ATIME_SET; + } + } + if (iattr->ia_valid & ATTR_MTIME) { + attrs->mask |= ORANGEFS_ATTR_SYS_MTIME; + if (iattr->ia_valid & ATTR_MTIME_SET) { + attrs->mtime = (time64_t)iattr->ia_mtime.tv_sec; + attrs->mask |= ORANGEFS_ATTR_SYS_MTIME_SET; + } + } + if (iattr->ia_valid & ATTR_CTIME) + attrs->mask |= ORANGEFS_ATTR_SYS_CTIME; + + /* + * ORANGEFS cannot set size with a setattr operation. Probably not likely + * to be requested through the VFS, but just in case, don't worry about + * ATTR_SIZE + */ + + if (iattr->ia_valid & ATTR_MODE) { + tmp_mode = iattr->ia_mode; + if (tmp_mode & (S_ISVTX)) { + if (is_root_handle(inode)) { + /* + * allow sticky bit to be set on root (since + * it shows up that way by default anyhow), + * but don't show it to the server + */ + tmp_mode -= S_ISVTX; + } else { + gossip_debug(GOSSIP_UTILS_DEBUG, + "User attempted to set sticky bit on non-root directory; returning EINVAL.\n"); + return -EINVAL; + } + } + + if (tmp_mode & (S_ISUID)) { + gossip_debug(GOSSIP_UTILS_DEBUG, + "Attempting to set setuid bit (not supported); returning EINVAL.\n"); + return -EINVAL; + } + + attrs->perms = ORANGEFS_util_translate_mode(tmp_mode); + attrs->mask |= ORANGEFS_ATTR_SYS_PERM; + } + + return 0; +} + +static int orangefs_inode_type(enum orangefs_ds_type objtype) +{ + if (objtype == ORANGEFS_TYPE_METAFILE) + return S_IFREG; + else if (objtype == ORANGEFS_TYPE_DIRECTORY) + return S_IFDIR; + else if (objtype == ORANGEFS_TYPE_SYMLINK) + return S_IFLNK; + else + return -1; +} + +static int orangefs_inode_is_stale(struct inode *inode, int new, + struct ORANGEFS_sys_attr_s *attrs, char *link_target) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + int type = orangefs_inode_type(attrs->objtype); + if (!new) { + /* + * If the inode type or symlink target have changed then this + * inode is stale. + */ + if (type == -1 || !(inode->i_mode & type)) { + orangefs_make_bad_inode(inode); + return 1; + } + if (type == S_IFLNK && strncmp(orangefs_inode->link_target, + link_target, ORANGEFS_NAME_MAX)) { + orangefs_make_bad_inode(inode); + return 1; + } + } + return 0; +} + +int orangefs_inode_getattr(struct inode *inode, int new, int size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + loff_t inode_size, rounded_up_size; + int ret, type; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: called on inode %pU\n", __func__, + get_khandle_from_ino(inode)); + + new_op = op_alloc(ORANGEFS_VFS_OP_GETATTR); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.getattr.refn = orangefs_inode->refn; + new_op->upcall.req.getattr.mask = size ? + ORANGEFS_ATTR_SYS_ALL_NOHINT : ORANGEFS_ATTR_SYS_ALL_NOHINT_NOSIZE; + + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + if (ret != 0) + goto out; + + type = orangefs_inode_type(new_op-> + downcall.resp.getattr.attributes.objtype); + ret = orangefs_inode_is_stale(inode, new, + &new_op->downcall.resp.getattr.attributes, + new_op->downcall.resp.getattr.link_target); + if (ret) { + ret = -ESTALE; + goto out; + } + + switch (type) { + case S_IFREG: + inode->i_flags = orangefs_inode_flags(&new_op-> + downcall.resp.getattr.attributes); + if (size) { + inode_size = (loff_t)new_op-> + downcall.resp.getattr.attributes.size; + rounded_up_size = + (inode_size + (4096 - (inode_size % 4096))); + inode->i_size = inode_size; + orangefs_inode->blksize = + new_op->downcall.resp.getattr.attributes.blksize; + spin_lock(&inode->i_lock); + inode->i_bytes = inode_size; + inode->i_blocks = + (unsigned long)(rounded_up_size / 512); + spin_unlock(&inode->i_lock); + } + break; + case S_IFDIR: + inode->i_size = PAGE_CACHE_SIZE; + orangefs_inode->blksize = (1 << inode->i_blkbits); + spin_lock(&inode->i_lock); + inode_set_bytes(inode, inode->i_size); + spin_unlock(&inode->i_lock); + set_nlink(inode, 1); + break; + case S_IFLNK: + if (new) { + inode->i_size = (loff_t)strlen(new_op-> + downcall.resp.getattr.link_target); + orangefs_inode->blksize = (1 << inode->i_blkbits); + strlcpy(orangefs_inode->link_target, + new_op->downcall.resp.getattr.link_target, + ORANGEFS_NAME_MAX); + inode->i_link = orangefs_inode->link_target; + } + break; + } + + inode->i_uid = make_kuid(&init_user_ns, new_op-> + downcall.resp.getattr.attributes.owner); + inode->i_gid = make_kgid(&init_user_ns, new_op-> + downcall.resp.getattr.attributes.group); + inode->i_atime.tv_sec = (time64_t)new_op-> + downcall.resp.getattr.attributes.atime; + inode->i_mtime.tv_sec = (time64_t)new_op-> + downcall.resp.getattr.attributes.mtime; + inode->i_ctime.tv_sec = (time64_t)new_op-> + downcall.resp.getattr.attributes.ctime; + inode->i_atime.tv_nsec = 0; + inode->i_mtime.tv_nsec = 0; + inode->i_ctime.tv_nsec = 0; + + /* special case: mark the root inode as sticky */ + inode->i_mode = type | (is_root_handle(inode) ? S_ISVTX : 0) | + orangefs_inode_perms(&new_op->downcall.resp.getattr.attributes); + + ret = 0; +out: + op_release(new_op); + return ret; +} + +int orangefs_inode_check_changed(struct inode *inode) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + int ret; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: called on inode %pU\n", __func__, + get_khandle_from_ino(inode)); + + new_op = op_alloc(ORANGEFS_VFS_OP_GETATTR); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.getattr.refn = orangefs_inode->refn; + new_op->upcall.req.getattr.mask = ORANGEFS_ATTR_SYS_TYPE | + ORANGEFS_ATTR_SYS_LNK_TARGET; + + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + if (ret != 0) + goto out; + + ret = orangefs_inode_is_stale(inode, 0, + &new_op->downcall.resp.getattr.attributes, + new_op->downcall.resp.getattr.link_target); +out: + op_release(new_op); + return ret; +} + +/* + * issues a orangefs setattr request to make sure the new attribute values + * take effect if successful. returns 0 on success; -errno otherwise + */ +int orangefs_inode_setattr(struct inode *inode, struct iattr *iattr) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + int ret; + + new_op = op_alloc(ORANGEFS_VFS_OP_SETATTR); + if (!new_op) + return -ENOMEM; + + new_op->upcall.req.setattr.refn = orangefs_inode->refn; + ret = copy_attributes_from_inode(inode, + &new_op->upcall.req.setattr.attributes, + iattr); + if (ret >= 0) { + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_inode_setattr: returning %d\n", + ret); + } + + op_release(new_op); + + /* + * successful setattr should clear the atime, mtime and + * ctime flags. + */ + if (ret == 0) { + ClearAtimeFlag(orangefs_inode); + ClearMtimeFlag(orangefs_inode); + ClearCtimeFlag(orangefs_inode); + ClearModeFlag(orangefs_inode); + } + + return ret; +} + +int orangefs_flush_inode(struct inode *inode) +{ + /* + * If it is a dirty inode, this function gets called. + * Gather all the information that needs to be setattr'ed + * Right now, this will only be used for mode, atime, mtime + * and/or ctime. + */ + struct iattr wbattr; + int ret; + int mtime_flag; + int ctime_flag; + int atime_flag; + int mode_flag; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + + memset(&wbattr, 0, sizeof(wbattr)); + + /* + * check inode flags up front, and clear them if they are set. This + * will prevent multiple processes from all trying to flush the same + * inode if they call close() simultaneously + */ + mtime_flag = MtimeFlag(orangefs_inode); + ClearMtimeFlag(orangefs_inode); + ctime_flag = CtimeFlag(orangefs_inode); + ClearCtimeFlag(orangefs_inode); + atime_flag = AtimeFlag(orangefs_inode); + ClearAtimeFlag(orangefs_inode); + mode_flag = ModeFlag(orangefs_inode); + ClearModeFlag(orangefs_inode); + + /* -- Lazy atime,mtime and ctime update -- + * Note: all times are dictated by server in the new scheme + * and not by the clients + * + * Also mode updates are being handled now.. + */ + + if (mtime_flag) + wbattr.ia_valid |= ATTR_MTIME; + if (ctime_flag) + wbattr.ia_valid |= ATTR_CTIME; + if (atime_flag) + wbattr.ia_valid |= ATTR_ATIME; + + if (mode_flag) { + wbattr.ia_mode = inode->i_mode; + wbattr.ia_valid |= ATTR_MODE; + } + + gossip_debug(GOSSIP_UTILS_DEBUG, + "*********** orangefs_flush_inode: %pU " + "(ia_valid %d)\n", + get_khandle_from_ino(inode), + wbattr.ia_valid); + if (wbattr.ia_valid == 0) { + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_flush_inode skipping setattr()\n"); + return 0; + } + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_flush_inode (%pU) writing mode %o\n", + get_khandle_from_ino(inode), + inode->i_mode); + + ret = orangefs_inode_setattr(inode, &wbattr); + + return ret; +} + +int orangefs_unmount_sb(struct super_block *sb) +{ + int ret = -EINVAL; + struct orangefs_kernel_op_s *new_op = NULL; + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_unmount_sb called on sb %p\n", + sb); + + new_op = op_alloc(ORANGEFS_VFS_OP_FS_UMOUNT); + if (!new_op) + return -ENOMEM; + new_op->upcall.req.fs_umount.id = ORANGEFS_SB(sb)->id; + new_op->upcall.req.fs_umount.fs_id = ORANGEFS_SB(sb)->fs_id; + strncpy(new_op->upcall.req.fs_umount.orangefs_config_server, + ORANGEFS_SB(sb)->devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + gossip_debug(GOSSIP_UTILS_DEBUG, + "Attempting ORANGEFS Unmount via host %s\n", + new_op->upcall.req.fs_umount.orangefs_config_server); + + ret = service_operation(new_op, "orangefs_fs_umount", 0); + + gossip_debug(GOSSIP_UTILS_DEBUG, + "orangefs_unmount: got return value of %d\n", ret); + if (ret) + sb = ERR_PTR(ret); + else + ORANGEFS_SB(sb)->mount_pending = 1; + + op_release(new_op); + return ret; +} + +void orangefs_make_bad_inode(struct inode *inode) +{ + if (is_root_handle(inode)) { + /* + * if this occurs, the pvfs2-client-core was killed but we + * can't afford to lose the inode operations and such + * associated with the root handle in any case. + */ + gossip_debug(GOSSIP_UTILS_DEBUG, + "*** NOT making bad root inode %pU\n", + get_khandle_from_ino(inode)); + } else { + gossip_debug(GOSSIP_UTILS_DEBUG, + "*** making bad inode %pU\n", + get_khandle_from_ino(inode)); + make_bad_inode(inode); + } +} + +/* + * The following is a very dirty hack that is now a permanent part of the + * ORANGEFS protocol. See protocol.h for more error definitions. + */ + +/* The order matches include/orangefs-types.h in the OrangeFS source. */ +static int PINT_errno_mapping[] = { + 0, EPERM, ENOENT, EINTR, EIO, ENXIO, EBADF, EAGAIN, ENOMEM, + EFAULT, EBUSY, EEXIST, ENODEV, ENOTDIR, EISDIR, EINVAL, EMFILE, + EFBIG, ENOSPC, EROFS, EMLINK, EPIPE, EDEADLK, ENAMETOOLONG, + ENOLCK, ENOSYS, ENOTEMPTY, ELOOP, EWOULDBLOCK, ENOMSG, EUNATCH, + EBADR, EDEADLOCK, ENODATA, ETIME, ENONET, EREMOTE, ECOMM, + EPROTO, EBADMSG, EOVERFLOW, ERESTART, EMSGSIZE, EPROTOTYPE, + ENOPROTOOPT, EPROTONOSUPPORT, EOPNOTSUPP, EADDRINUSE, + EADDRNOTAVAIL, ENETDOWN, ENETUNREACH, ENETRESET, ENOBUFS, + ETIMEDOUT, ECONNREFUSED, EHOSTDOWN, EHOSTUNREACH, EALREADY, + EACCES, ECONNRESET, ERANGE +}; + +int orangefs_normalize_to_errno(__s32 error_code) +{ + __u32 i; + + /* Success */ + if (error_code == 0) { + return 0; + /* + * This shouldn't ever happen. If it does it should be fixed on the + * server. + */ + } else if (error_code > 0) { + gossip_err("orangefs: error status receieved.\n"); + gossip_err("orangefs: assuming error code is inverted.\n"); + error_code = -error_code; + } + + /* + * XXX: This is very bad since error codes from ORANGEFS may not be + * suitable for return into userspace. + */ + + /* + * Convert ORANGEFS error values into errno values suitable for return + * from the kernel. + */ + if ((-error_code) & ORANGEFS_NON_ERRNO_ERROR_BIT) { + if (((-error_code) & + (ORANGEFS_ERROR_NUMBER_BITS|ORANGEFS_NON_ERRNO_ERROR_BIT| + ORANGEFS_ERROR_BIT)) == ORANGEFS_ECANCEL) { + /* + * cancellation error codes generally correspond to + * a timeout from the client's perspective + */ + error_code = -ETIMEDOUT; + } else { + /* assume a default error code */ + gossip_err("orangefs: warning: got error code without errno equivalent: %d.\n", error_code); + error_code = -EINVAL; + } + + /* Convert ORANGEFS encoded errno values into regular errno values. */ + } else if ((-error_code) & ORANGEFS_ERROR_BIT) { + i = (-error_code) & ~(ORANGEFS_ERROR_BIT|ORANGEFS_ERROR_CLASS_BITS); + if (i < sizeof(PINT_errno_mapping)/sizeof(*PINT_errno_mapping)) + error_code = -PINT_errno_mapping[i]; + else + error_code = -EINVAL; + + /* + * Only ORANGEFS protocol error codes should ever come here. Otherwise + * there is a bug somewhere. + */ + } else { + gossip_err("orangefs: orangefs_normalize_to_errno: got error code which is not from ORANGEFS.\n"); + } + return error_code; +} + +#define NUM_MODES 11 +__s32 ORANGEFS_util_translate_mode(int mode) +{ + int ret = 0; + int i = 0; + static int modes[NUM_MODES] = { + S_IXOTH, S_IWOTH, S_IROTH, + S_IXGRP, S_IWGRP, S_IRGRP, + S_IXUSR, S_IWUSR, S_IRUSR, + S_ISGID, S_ISUID + }; + static int orangefs_modes[NUM_MODES] = { + ORANGEFS_O_EXECUTE, ORANGEFS_O_WRITE, ORANGEFS_O_READ, + ORANGEFS_G_EXECUTE, ORANGEFS_G_WRITE, ORANGEFS_G_READ, + ORANGEFS_U_EXECUTE, ORANGEFS_U_WRITE, ORANGEFS_U_READ, + ORANGEFS_G_SGID, ORANGEFS_U_SUID + }; + + for (i = 0; i < NUM_MODES; i++) + if (mode & modes[i]) + ret |= orangefs_modes[i]; + + return ret; +} +#undef NUM_MODES + +/* + * After obtaining a string representation of the client's debug + * keywords and their associated masks, this function is called to build an + * array of these values. + */ +int orangefs_prepare_cdm_array(char *debug_array_string) +{ + int i; + int rc = -EINVAL; + char *cds_head = NULL; + char *cds_delimiter = NULL; + int keyword_len = 0; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + /* + * figure out how many elements the cdm_array needs. + */ + for (i = 0; i < strlen(debug_array_string); i++) + if (debug_array_string[i] == '\n') + cdm_element_count++; + + if (!cdm_element_count) { + pr_info("No elements in client debug array string!\n"); + goto out; + } + + cdm_array = + kzalloc(cdm_element_count * sizeof(struct client_debug_mask), + GFP_KERNEL); + if (!cdm_array) { + pr_info("malloc failed for cdm_array!\n"); + rc = -ENOMEM; + goto out; + } + + cds_head = debug_array_string; + + for (i = 0; i < cdm_element_count; i++) { + cds_delimiter = strchr(cds_head, '\n'); + *cds_delimiter = '\0'; + + keyword_len = strcspn(cds_head, " "); + + cdm_array[i].keyword = kzalloc(keyword_len + 1, GFP_KERNEL); + if (!cdm_array[i].keyword) { + rc = -ENOMEM; + goto out; + } + + sscanf(cds_head, + "%s %llx %llx", + cdm_array[i].keyword, + (unsigned long long *)&(cdm_array[i].mask1), + (unsigned long long *)&(cdm_array[i].mask2)); + + if (!strcmp(cdm_array[i].keyword, ORANGEFS_VERBOSE)) + client_verbose_index = i; + + if (!strcmp(cdm_array[i].keyword, ORANGEFS_ALL)) + client_all_index = i; + + cds_head = cds_delimiter + 1; + } + + rc = cdm_element_count; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: rc:%d:\n", __func__, rc); + +out: + + return rc; + +} + +/* + * /sys/kernel/debug/orangefs/debug-help can be catted to + * see all the available kernel and client debug keywords. + * + * When the kernel boots, we have no idea what keywords the + * client supports, nor their associated masks. + * + * We pass through this function once at boot and stamp a + * boilerplate "we don't know" message for the client in the + * debug-help file. We pass through here again when the client + * starts and then we can fill out the debug-help file fully. + * + * The client might be restarted any number of times between + * reboots, we only build the debug-help file the first time. + */ +int orangefs_prepare_debugfs_help_string(int at_boot) +{ + int rc = -EINVAL; + int i; + int byte_count = 0; + char *client_title = "Client Debug Keywords:\n"; + char *kernel_title = "Kernel Debug Keywords:\n"; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + if (at_boot) { + byte_count += strlen(HELP_STRING_UNINITIALIZED); + client_title = HELP_STRING_UNINITIALIZED; + } else { + /* + * fill the client keyword/mask array and remember + * how many elements there were. + */ + cdm_element_count = + orangefs_prepare_cdm_array(client_debug_array_string); + if (cdm_element_count <= 0) + goto out; + + /* Count the bytes destined for debug_help_string. */ + byte_count += strlen(client_title); + + for (i = 0; i < cdm_element_count; i++) { + byte_count += strlen(cdm_array[i].keyword + 2); + if (byte_count >= DEBUG_HELP_STRING_SIZE) { + pr_info("%s: overflow 1!\n", __func__); + goto out; + } + } + + gossip_debug(GOSSIP_UTILS_DEBUG, + "%s: cdm_element_count:%d:\n", + __func__, + cdm_element_count); + } + + byte_count += strlen(kernel_title); + for (i = 0; i < num_kmod_keyword_mask_map; i++) { + byte_count += + strlen(s_kmod_keyword_mask_map[i].keyword + 2); + if (byte_count >= DEBUG_HELP_STRING_SIZE) { + pr_info("%s: overflow 2!\n", __func__); + goto out; + } + } + + /* build debug_help_string. */ + debug_help_string = kzalloc(DEBUG_HELP_STRING_SIZE, GFP_KERNEL); + if (!debug_help_string) { + rc = -ENOMEM; + goto out; + } + + strcat(debug_help_string, client_title); + + if (!at_boot) { + for (i = 0; i < cdm_element_count; i++) { + strcat(debug_help_string, "\t"); + strcat(debug_help_string, cdm_array[i].keyword); + strcat(debug_help_string, "\n"); + } + } + + strcat(debug_help_string, "\n"); + strcat(debug_help_string, kernel_title); + + for (i = 0; i < num_kmod_keyword_mask_map; i++) { + strcat(debug_help_string, "\t"); + strcat(debug_help_string, s_kmod_keyword_mask_map[i].keyword); + strcat(debug_help_string, "\n"); + } + + rc = 0; + +out: + + return rc; + +} + +/* + * kernel = type 0 + * client = type 1 + */ +void debug_mask_to_string(void *mask, int type) +{ + int i; + int len = 0; + char *debug_string; + int element_count = 0; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + if (type) { + debug_string = client_debug_string; + element_count = cdm_element_count; + } else { + debug_string = kernel_debug_string; + element_count = num_kmod_keyword_mask_map; + } + + memset(debug_string, 0, ORANGEFS_MAX_DEBUG_STRING_LEN); + + /* + * Some keywords, like "all" or "verbose", are amalgams of + * numerous other keywords. Make a special check for those + * before grinding through the whole mask only to find out + * later... + */ + if (check_amalgam_keyword(mask, type)) + goto out; + + /* Build the debug string. */ + for (i = 0; i < element_count; i++) + if (type) + do_c_string(mask, i); + else + do_k_string(mask, i); + + len = strlen(debug_string); + + if ((len) && (type)) + client_debug_string[len - 1] = '\0'; + else if (len) + kernel_debug_string[len - 1] = '\0'; + else if (type) + strcpy(client_debug_string, "none"); + else + strcpy(kernel_debug_string, "none"); + +out: +gossip_debug(GOSSIP_UTILS_DEBUG, "%s: string:%s:\n", __func__, debug_string); + + return; + +} + +void do_k_string(void *k_mask, int index) +{ + __u64 *mask = (__u64 *) k_mask; + + if (keyword_is_amalgam((char *) s_kmod_keyword_mask_map[index].keyword)) + goto out; + + if (*mask & s_kmod_keyword_mask_map[index].mask_val) { + if ((strlen(kernel_debug_string) + + strlen(s_kmod_keyword_mask_map[index].keyword)) + < ORANGEFS_MAX_DEBUG_STRING_LEN - 1) { + strcat(kernel_debug_string, + s_kmod_keyword_mask_map[index].keyword); + strcat(kernel_debug_string, ","); + } else { + gossip_err("%s: overflow!\n", __func__); + strcpy(kernel_debug_string, ORANGEFS_ALL); + goto out; + } + } + +out: + + return; +} + +void do_c_string(void *c_mask, int index) +{ + struct client_debug_mask *mask = (struct client_debug_mask *) c_mask; + + if (keyword_is_amalgam(cdm_array[index].keyword)) + goto out; + + if ((mask->mask1 & cdm_array[index].mask1) || + (mask->mask2 & cdm_array[index].mask2)) { + if ((strlen(client_debug_string) + + strlen(cdm_array[index].keyword) + 1) + < ORANGEFS_MAX_DEBUG_STRING_LEN - 2) { + strcat(client_debug_string, + cdm_array[index].keyword); + strcat(client_debug_string, ","); + } else { + gossip_err("%s: overflow!\n", __func__); + strcpy(client_debug_string, ORANGEFS_ALL); + goto out; + } + } +out: + return; +} + +int keyword_is_amalgam(char *keyword) +{ + int rc = 0; + + if ((!strcmp(keyword, ORANGEFS_ALL)) || (!strcmp(keyword, ORANGEFS_VERBOSE))) + rc = 1; + + return rc; +} + +/* + * kernel = type 0 + * client = type 1 + * + * return 1 if we found an amalgam. + */ +int check_amalgam_keyword(void *mask, int type) +{ + __u64 *k_mask; + struct client_debug_mask *c_mask; + int k_all_index = num_kmod_keyword_mask_map - 1; + int rc = 0; + + if (type) { + c_mask = (struct client_debug_mask *) mask; + + if ((c_mask->mask1 == cdm_array[client_all_index].mask1) && + (c_mask->mask2 == cdm_array[client_all_index].mask2)) { + strcpy(client_debug_string, ORANGEFS_ALL); + rc = 1; + goto out; + } + + if ((c_mask->mask1 == cdm_array[client_verbose_index].mask1) && + (c_mask->mask2 == cdm_array[client_verbose_index].mask2)) { + strcpy(client_debug_string, ORANGEFS_VERBOSE); + rc = 1; + goto out; + } + + } else { + k_mask = (__u64 *) mask; + + if (*k_mask >= s_kmod_keyword_mask_map[k_all_index].mask_val) { + strcpy(kernel_debug_string, ORANGEFS_ALL); + rc = 1; + goto out; + } + } + +out: + + return rc; +} + +/* + * kernel = type 0 + * client = type 1 + */ +void debug_string_to_mask(char *debug_string, void *mask, int type) +{ + char *unchecked_keyword; + int i; + char *strsep_fodder = kstrdup(debug_string, GFP_KERNEL); + char *original_pointer; + int element_count = 0; + struct client_debug_mask *c_mask; + __u64 *k_mask; + + gossip_debug(GOSSIP_UTILS_DEBUG, "%s: start\n", __func__); + + if (type) { + c_mask = (struct client_debug_mask *)mask; + element_count = cdm_element_count; + } else { + k_mask = (__u64 *)mask; + *k_mask = 0; + element_count = num_kmod_keyword_mask_map; + } + + original_pointer = strsep_fodder; + while ((unchecked_keyword = strsep(&strsep_fodder, ","))) + if (strlen(unchecked_keyword)) { + for (i = 0; i < element_count; i++) + if (type) + do_c_mask(i, + unchecked_keyword, + &c_mask); + else + do_k_mask(i, + unchecked_keyword, + &k_mask); + } + + kfree(original_pointer); +} + +void do_c_mask(int i, + char *unchecked_keyword, + struct client_debug_mask **sane_mask) +{ + + if (!strcmp(cdm_array[i].keyword, unchecked_keyword)) { + (**sane_mask).mask1 = (**sane_mask).mask1 | cdm_array[i].mask1; + (**sane_mask).mask2 = (**sane_mask).mask2 | cdm_array[i].mask2; + } +} + +void do_k_mask(int i, char *unchecked_keyword, __u64 **sane_mask) +{ + + if (!strcmp(s_kmod_keyword_mask_map[i].keyword, unchecked_keyword)) + **sane_mask = (**sane_mask) | + s_kmod_keyword_mask_map[i].mask_val; +} diff --git a/fs/orangefs/protocol.h b/fs/orangefs/protocol.h new file mode 100644 index 000000000000..50578a28bd9e --- /dev/null +++ b/fs/orangefs/protocol.h @@ -0,0 +1,452 @@ +#include <linux/types.h> +#include <linux/spinlock_types.h> +#include <linux/slab.h> +#include <linux/ioctl.h> + +extern struct client_debug_mask *cdm_array; +extern char *debug_help_string; +extern int help_string_initialized; +extern struct dentry *debug_dir; +extern struct dentry *help_file_dentry; +extern struct dentry *client_debug_dentry; +extern const struct file_operations debug_help_fops; +extern int client_all_index; +extern int client_verbose_index; +extern int cdm_element_count; +#define DEBUG_HELP_STRING_SIZE 4096 +#define HELP_STRING_UNINITIALIZED \ + "Client Debug Keywords are unknown until the first time\n" \ + "the client is started after boot.\n" +#define ORANGEFS_KMOD_DEBUG_HELP_FILE "debug-help" +#define ORANGEFS_KMOD_DEBUG_FILE "kernel-debug" +#define ORANGEFS_CLIENT_DEBUG_FILE "client-debug" +#define ORANGEFS_VERBOSE "verbose" +#define ORANGEFS_ALL "all" + +/* pvfs2-config.h ***********************************************************/ +#define ORANGEFS_VERSION_MAJOR 2 +#define ORANGEFS_VERSION_MINOR 9 +#define ORANGEFS_VERSION_SUB 0 + +/* khandle stuff ***********************************************************/ + +/* + * The 2.9 core will put 64 bit handles in here like this: + * 1234 0000 0000 5678 + * The 3.0 and beyond cores will put 128 bit handles in here like this: + * 1234 5678 90AB CDEF + * The kernel module will always use the first four bytes and + * the last four bytes as an inum. + */ +struct orangefs_khandle { + unsigned char u[16]; +} __aligned(8); + +/* + * kernel version of an object ref. + */ +struct orangefs_object_kref { + struct orangefs_khandle khandle; + __s32 fs_id; + __s32 __pad1; +}; + +/* + * compare 2 khandles assumes little endian thus from large address to + * small address + */ +static inline int ORANGEFS_khandle_cmp(const struct orangefs_khandle *kh1, + const struct orangefs_khandle *kh2) +{ + int i; + + for (i = 15; i >= 0; i--) { + if (kh1->u[i] > kh2->u[i]) + return 1; + if (kh1->u[i] < kh2->u[i]) + return -1; + } + + return 0; +} + +static inline void ORANGEFS_khandle_to(const struct orangefs_khandle *kh, + void *p, int size) +{ + + memset(p, 0, size); + memcpy(p, kh->u, 16); + +} + +static inline void ORANGEFS_khandle_from(struct orangefs_khandle *kh, + void *p, int size) +{ + memset(kh, 0, 16); + memcpy(kh->u, p, 16); + +} + +/* pvfs2-types.h ************************************************************/ +typedef __u32 ORANGEFS_uid; +typedef __u32 ORANGEFS_gid; +typedef __s32 ORANGEFS_fs_id; +typedef __u32 ORANGEFS_permissions; +typedef __u64 ORANGEFS_time; +typedef __s64 ORANGEFS_size; +typedef __u64 ORANGEFS_flags; +typedef __u64 ORANGEFS_ds_position; +typedef __s32 ORANGEFS_error; +typedef __s64 ORANGEFS_offset; + +#define ORANGEFS_SUPER_MAGIC 0x20030528 + +/* + * ORANGEFS error codes are a signed 32-bit integer. Error codes are negative, but + * the sign is stripped before decoding. + */ + +/* Bit 31 is not used since it is the sign. */ + +/* + * Bit 30 specifies that this is a ORANGEFS error. A ORANGEFS error is either an + * encoded errno value or a ORANGEFS protocol error. + */ +#define ORANGEFS_ERROR_BIT (1 << 30) + +/* + * Bit 29 specifies that this is a ORANGEFS protocol error and not an encoded + * errno value. + */ +#define ORANGEFS_NON_ERRNO_ERROR_BIT (1 << 29) + +/* + * Bits 9, 8, and 7 specify the error class, which encodes the section of + * server code the error originated in for logging purposes. It is not used + * in the kernel except to be masked out. + */ +#define ORANGEFS_ERROR_CLASS_BITS 0x380 + +/* Bits 6 - 0 are reserved for the actual error code. */ +#define ORANGEFS_ERROR_NUMBER_BITS 0x7f + +/* Encoded errno values decoded by PINT_errno_mapping in orangefs-utils.c. */ + +/* Our own ORANGEFS protocol error codes. */ +#define ORANGEFS_ECANCEL (1|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EDEVINIT (2|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EDETAIL (3|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EHOSTNTFD (4|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_EADDRNTFD (5|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ENORECVR (6|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ETRYAGAIN (7|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ENOTPVFS (8|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) +#define ORANGEFS_ESECURITY (9|ORANGEFS_NON_ERRNO_ERROR_BIT|ORANGEFS_ERROR_BIT) + +/* permission bits */ +#define ORANGEFS_O_EXECUTE (1 << 0) +#define ORANGEFS_O_WRITE (1 << 1) +#define ORANGEFS_O_READ (1 << 2) +#define ORANGEFS_G_EXECUTE (1 << 3) +#define ORANGEFS_G_WRITE (1 << 4) +#define ORANGEFS_G_READ (1 << 5) +#define ORANGEFS_U_EXECUTE (1 << 6) +#define ORANGEFS_U_WRITE (1 << 7) +#define ORANGEFS_U_READ (1 << 8) +/* no ORANGEFS_U_VTX (sticky bit) */ +#define ORANGEFS_G_SGID (1 << 10) +#define ORANGEFS_U_SUID (1 << 11) + +/* definition taken from stdint.h */ +#define INT32_MAX (2147483647) +#define ORANGEFS_ITERATE_START (INT32_MAX - 1) +#define ORANGEFS_ITERATE_END (INT32_MAX - 2) +#define ORANGEFS_ITERATE_NEXT (INT32_MAX - 3) +#define ORANGEFS_READDIR_START ORANGEFS_ITERATE_START +#define ORANGEFS_READDIR_END ORANGEFS_ITERATE_END +#define ORANGEFS_IMMUTABLE_FL FS_IMMUTABLE_FL +#define ORANGEFS_APPEND_FL FS_APPEND_FL +#define ORANGEFS_NOATIME_FL FS_NOATIME_FL +#define ORANGEFS_MIRROR_FL 0x01000000ULL +#define ORANGEFS_O_EXECUTE (1 << 0) +#define ORANGEFS_FS_ID_NULL ((__s32)0) +#define ORANGEFS_ATTR_SYS_UID (1 << 0) +#define ORANGEFS_ATTR_SYS_GID (1 << 1) +#define ORANGEFS_ATTR_SYS_PERM (1 << 2) +#define ORANGEFS_ATTR_SYS_ATIME (1 << 3) +#define ORANGEFS_ATTR_SYS_CTIME (1 << 4) +#define ORANGEFS_ATTR_SYS_MTIME (1 << 5) +#define ORANGEFS_ATTR_SYS_TYPE (1 << 6) +#define ORANGEFS_ATTR_SYS_ATIME_SET (1 << 7) +#define ORANGEFS_ATTR_SYS_MTIME_SET (1 << 8) +#define ORANGEFS_ATTR_SYS_SIZE (1 << 20) +#define ORANGEFS_ATTR_SYS_LNK_TARGET (1 << 24) +#define ORANGEFS_ATTR_SYS_DFILE_COUNT (1 << 25) +#define ORANGEFS_ATTR_SYS_DIRENT_COUNT (1 << 26) +#define ORANGEFS_ATTR_SYS_BLKSIZE (1 << 28) +#define ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT (1 << 29) +#define ORANGEFS_ATTR_SYS_COMMON_ALL \ + (ORANGEFS_ATTR_SYS_UID | \ + ORANGEFS_ATTR_SYS_GID | \ + ORANGEFS_ATTR_SYS_PERM | \ + ORANGEFS_ATTR_SYS_ATIME | \ + ORANGEFS_ATTR_SYS_CTIME | \ + ORANGEFS_ATTR_SYS_MTIME | \ + ORANGEFS_ATTR_SYS_TYPE) + +#define ORANGEFS_ATTR_SYS_ALL_SETABLE \ +(ORANGEFS_ATTR_SYS_COMMON_ALL-ORANGEFS_ATTR_SYS_TYPE) + +#define ORANGEFS_ATTR_SYS_ALL_NOHINT \ + (ORANGEFS_ATTR_SYS_COMMON_ALL | \ + ORANGEFS_ATTR_SYS_SIZE | \ + ORANGEFS_ATTR_SYS_LNK_TARGET | \ + ORANGEFS_ATTR_SYS_DFILE_COUNT | \ + ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT | \ + ORANGEFS_ATTR_SYS_DIRENT_COUNT | \ + ORANGEFS_ATTR_SYS_BLKSIZE) + +#define ORANGEFS_ATTR_SYS_ALL_NOHINT_NOSIZE \ + (ORANGEFS_ATTR_SYS_COMMON_ALL | \ + ORANGEFS_ATTR_SYS_LNK_TARGET | \ + ORANGEFS_ATTR_SYS_DFILE_COUNT | \ + ORANGEFS_ATTR_SYS_MIRROR_COPIES_COUNT | \ + ORANGEFS_ATTR_SYS_DIRENT_COUNT | \ + ORANGEFS_ATTR_SYS_BLKSIZE) + +#define ORANGEFS_XATTR_REPLACE 0x2 +#define ORANGEFS_XATTR_CREATE 0x1 +#define ORANGEFS_MAX_SERVER_ADDR_LEN 256 +#define ORANGEFS_NAME_MAX 256 +/* + * max extended attribute name len as imposed by the VFS and exploited for the + * upcall request types. + * NOTE: Please retain them as multiples of 8 even if you wish to change them + * This is *NECESSARY* for supporting 32 bit user-space binaries on a 64-bit + * kernel. Due to implementation within DBPF, this really needs to be + * ORANGEFS_NAME_MAX, which it was the same value as, but no reason to let it + * break if that changes in the future. + */ +#define ORANGEFS_MAX_XATTR_NAMELEN ORANGEFS_NAME_MAX /* Not the same as + * XATTR_NAME_MAX defined + * by <linux/xattr.h> + */ +#define ORANGEFS_MAX_XATTR_VALUELEN 8192 /* Not the same as XATTR_SIZE_MAX + * defined by <linux/xattr.h> + */ +#define ORANGEFS_MAX_XATTR_LISTLEN 16 /* Not the same as XATTR_LIST_MAX + * defined by <linux/xattr.h> + */ +/* + * ORANGEFS I/O operation types, used in both system and server interfaces. + */ +enum ORANGEFS_io_type { + ORANGEFS_IO_READ = 1, + ORANGEFS_IO_WRITE = 2 +}; + +/* + * If this enum is modified the server parameters related to the precreate pool + * batch and low threshold sizes may need to be modified to reflect this + * change. + */ +enum orangefs_ds_type { + ORANGEFS_TYPE_NONE = 0, + ORANGEFS_TYPE_METAFILE = (1 << 0), + ORANGEFS_TYPE_DATAFILE = (1 << 1), + ORANGEFS_TYPE_DIRECTORY = (1 << 2), + ORANGEFS_TYPE_SYMLINK = (1 << 3), + ORANGEFS_TYPE_DIRDATA = (1 << 4), + ORANGEFS_TYPE_INTERNAL = (1 << 5) /* for the server's private use */ +}; + +/* + * ORANGEFS_certificate simply stores a buffer with the buffer size. + * The buffer can be converted to an OpenSSL X509 struct for use. + */ +struct ORANGEFS_certificate { + __u32 buf_size; + unsigned char *buf; +}; + +/* + * A credential identifies a user and is signed by the client/user + * private key. + */ +struct ORANGEFS_credential { + __u32 userid; /* user id */ + __u32 num_groups; /* length of group_array */ + __u32 *group_array; /* groups for which the user is a member */ + char *issuer; /* alias of the issuing server */ + __u64 timeout; /* seconds after epoch to time out */ + __u32 sig_size; /* length of the signature in bytes */ + unsigned char *signature; /* digital signature */ + struct ORANGEFS_certificate certificate; /* user certificate buffer */ +}; +#define extra_size_ORANGEFS_credential (ORANGEFS_REQ_LIMIT_GROUPS * \ + sizeof(__u32) + \ + ORANGEFS_REQ_LIMIT_ISSUER + \ + ORANGEFS_REQ_LIMIT_SIGNATURE + \ + extra_size_ORANGEFS_certificate) + +/* This structure is used by the VFS-client interaction alone */ +struct ORANGEFS_keyval_pair { + char key[ORANGEFS_MAX_XATTR_NAMELEN]; + __s32 key_sz; /* __s32 for portable, fixed-size structures */ + __s32 val_sz; + char val[ORANGEFS_MAX_XATTR_VALUELEN]; +}; + +/* pvfs2-sysint.h ***********************************************************/ +/* Describes attributes for a file, directory, or symlink. */ +struct ORANGEFS_sys_attr_s { + __u32 owner; + __u32 group; + __u32 perms; + __u64 atime; + __u64 mtime; + __u64 ctime; + __s64 size; + + /* NOTE: caller must free if valid */ + char *link_target; + + /* Changed to __s32 so that size of structure does not change */ + __s32 dfile_count; + + /* Changed to __s32 so that size of structure does not change */ + __s32 distr_dir_servers_initial; + + /* Changed to __s32 so that size of structure does not change */ + __s32 distr_dir_servers_max; + + /* Changed to __s32 so that size of structure does not change */ + __s32 distr_dir_split_size; + + __u32 mirror_copies_count; + + /* NOTE: caller must free if valid */ + char *dist_name; + + /* NOTE: caller must free if valid */ + char *dist_params; + + __s64 dirent_count; + enum orangefs_ds_type objtype; + __u64 flags; + __u32 mask; + __s64 blksize; +}; + +#define ORANGEFS_LOOKUP_LINK_NO_FOLLOW 0 + +/* pint-dev.h ***************************************************************/ + +/* parameter structure used in ORANGEFS_DEV_DEBUG ioctl command */ +struct dev_mask_info_s { + enum { + KERNEL_MASK, + CLIENT_MASK, + } mask_type; + __u64 mask_value; +}; + +struct dev_mask2_info_s { + __u64 mask1_value; + __u64 mask2_value; +}; + +/* pvfs2-util.h *************************************************************/ +__s32 ORANGEFS_util_translate_mode(int mode); + +/* pvfs2-debug.h ************************************************************/ +#include "orangefs-debug.h" + +/* pvfs2-internal.h *********************************************************/ +#define llu(x) (unsigned long long)(x) +#define lld(x) (long long)(x) + +/* pint-dev-shared.h ********************************************************/ +#define ORANGEFS_DEV_MAGIC 'k' + +#define ORANGEFS_READDIR_DEFAULT_DESC_COUNT 5 + +#define DEV_GET_MAGIC 0x1 +#define DEV_GET_MAX_UPSIZE 0x2 +#define DEV_GET_MAX_DOWNSIZE 0x3 +#define DEV_MAP 0x4 +#define DEV_REMOUNT_ALL 0x5 +#define DEV_DEBUG 0x6 +#define DEV_UPSTREAM 0x7 +#define DEV_CLIENT_MASK 0x8 +#define DEV_CLIENT_STRING 0x9 +#define DEV_MAX_NR 0xa + +/* supported ioctls, codes are with respect to user-space */ +enum { + ORANGEFS_DEV_GET_MAGIC = _IOW(ORANGEFS_DEV_MAGIC, DEV_GET_MAGIC, __s32), + ORANGEFS_DEV_GET_MAX_UPSIZE = + _IOW(ORANGEFS_DEV_MAGIC, DEV_GET_MAX_UPSIZE, __s32), + ORANGEFS_DEV_GET_MAX_DOWNSIZE = + _IOW(ORANGEFS_DEV_MAGIC, DEV_GET_MAX_DOWNSIZE, __s32), + ORANGEFS_DEV_MAP = _IO(ORANGEFS_DEV_MAGIC, DEV_MAP), + ORANGEFS_DEV_REMOUNT_ALL = _IO(ORANGEFS_DEV_MAGIC, DEV_REMOUNT_ALL), + ORANGEFS_DEV_DEBUG = _IOR(ORANGEFS_DEV_MAGIC, DEV_DEBUG, __s32), + ORANGEFS_DEV_UPSTREAM = _IOW(ORANGEFS_DEV_MAGIC, DEV_UPSTREAM, int), + ORANGEFS_DEV_CLIENT_MASK = _IOW(ORANGEFS_DEV_MAGIC, + DEV_CLIENT_MASK, + struct dev_mask2_info_s), + ORANGEFS_DEV_CLIENT_STRING = _IOW(ORANGEFS_DEV_MAGIC, + DEV_CLIENT_STRING, + char *), + ORANGEFS_DEV_MAXNR = DEV_MAX_NR, +}; + +/* + * version number for use in communicating between kernel space and user + * space. Zero signifies the upstream version of the kernel module. + */ +#define ORANGEFS_KERNEL_PROTO_VERSION 0 +#define ORANGEFS_MINIMUM_USERSPACE_VERSION 20903 + +/* + * describes memory regions to map in the ORANGEFS_DEV_MAP ioctl. + * NOTE: See devorangefs-req.c for 32 bit compat structure. + * Since this structure has a variable-sized layout that is different + * on 32 and 64 bit platforms, we need to normalize to a 64 bit layout + * on such systems before servicing ioctl calls from user-space binaries + * that may be 32 bit! + */ +struct ORANGEFS_dev_map_desc { + void *ptr; + __s32 total_size; + __s32 size; + __s32 count; +}; + +/* gossip.h *****************************************************************/ + +#ifdef GOSSIP_DISABLE_DEBUG +#define gossip_debug(mask, format, f...) do {} while (0) +#else +extern __u64 gossip_debug_mask; +extern struct client_debug_mask client_debug_mask; + +/* try to avoid function call overhead by checking masks in macro */ +#define gossip_debug(mask, format, f...) \ +do { \ + if (gossip_debug_mask & mask) \ + printk(format, ##f); \ +} while (0) +#endif /* GOSSIP_DISABLE_DEBUG */ + +/* do file and line number printouts w/ the GNU preprocessor */ +#define gossip_ldebug(mask, format, f...) \ + gossip_debug(mask, "%s: " format, __func__, ##f) + +#define gossip_err printk +#define gossip_lerr(format, f...) \ + gossip_err("%s line %d: " format, \ + __FILE__, \ + __LINE__, \ + ##f) diff --git a/fs/orangefs/super.c b/fs/orangefs/super.c new file mode 100644 index 000000000000..b9da9a0281c9 --- /dev/null +++ b/fs/orangefs/super.c @@ -0,0 +1,559 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +#include <linux/parser.h> + +/* a cache for orangefs-inode objects (i.e. orangefs inode private data) */ +static struct kmem_cache *orangefs_inode_cache; + +/* list for storing orangefs specific superblocks in use */ +LIST_HEAD(orangefs_superblocks); + +DEFINE_SPINLOCK(orangefs_superblocks_lock); + +enum { + Opt_intr, + Opt_acl, + Opt_local_lock, + + Opt_err +}; + +static const match_table_t tokens = { + { Opt_acl, "acl" }, + { Opt_intr, "intr" }, + { Opt_local_lock, "local_lock" }, + { Opt_err, NULL } +}; + + +static int parse_mount_options(struct super_block *sb, char *options, + int silent) +{ + struct orangefs_sb_info_s *orangefs_sb = ORANGEFS_SB(sb); + substring_t args[MAX_OPT_ARGS]; + char *p; + + /* + * Force any potential flags that might be set from the mount + * to zero, ie, initialize to unset. + */ + sb->s_flags &= ~MS_POSIXACL; + orangefs_sb->flags &= ~ORANGEFS_OPT_INTR; + orangefs_sb->flags &= ~ORANGEFS_OPT_LOCAL_LOCK; + + while ((p = strsep(&options, ",")) != NULL) { + int token; + + if (!*p) + continue; + + token = match_token(p, tokens, args); + switch (token) { + case Opt_acl: + sb->s_flags |= MS_POSIXACL; + break; + case Opt_intr: + orangefs_sb->flags |= ORANGEFS_OPT_INTR; + break; + case Opt_local_lock: + orangefs_sb->flags |= ORANGEFS_OPT_LOCAL_LOCK; + break; + default: + goto fail; + } + } + + return 0; +fail: + if (!silent) + gossip_err("Error: mount option [%s] is not supported.\n", p); + return -EINVAL; +} + +static void orangefs_inode_cache_ctor(void *req) +{ + struct orangefs_inode_s *orangefs_inode = req; + + inode_init_once(&orangefs_inode->vfs_inode); + init_rwsem(&orangefs_inode->xattr_sem); + + orangefs_inode->vfs_inode.i_version = 1; +} + +static struct inode *orangefs_alloc_inode(struct super_block *sb) +{ + struct orangefs_inode_s *orangefs_inode; + + orangefs_inode = kmem_cache_alloc(orangefs_inode_cache, GFP_KERNEL); + if (orangefs_inode == NULL) { + gossip_err("Failed to allocate orangefs_inode\n"); + return NULL; + } + + /* + * We want to clear everything except for rw_semaphore and the + * vfs_inode. + */ + memset(&orangefs_inode->refn.khandle, 0, 16); + orangefs_inode->refn.fs_id = ORANGEFS_FS_ID_NULL; + orangefs_inode->last_failed_block_index_read = 0; + memset(orangefs_inode->link_target, 0, sizeof(orangefs_inode->link_target)); + orangefs_inode->pinode_flags = 0; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_alloc_inode: allocated %p\n", + &orangefs_inode->vfs_inode); + return &orangefs_inode->vfs_inode; +} + +static void orangefs_destroy_inode(struct inode *inode) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "%s: deallocated %p destroying inode %pU\n", + __func__, orangefs_inode, get_khandle_from_ino(inode)); + + kmem_cache_free(orangefs_inode_cache, orangefs_inode); +} + +/* + * NOTE: information filled in here is typically reflected in the + * output of the system command 'df' +*/ +static int orangefs_statfs(struct dentry *dentry, struct kstatfs *buf) +{ + int ret = -ENOMEM; + struct orangefs_kernel_op_s *new_op = NULL; + int flags = 0; + struct super_block *sb = NULL; + + sb = dentry->d_sb; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_statfs: called on sb %p (fs_id is %d)\n", + sb, + (int)(ORANGEFS_SB(sb)->fs_id)); + + new_op = op_alloc(ORANGEFS_VFS_OP_STATFS); + if (!new_op) + return ret; + new_op->upcall.req.statfs.fs_id = ORANGEFS_SB(sb)->fs_id; + + if (ORANGEFS_SB(sb)->flags & ORANGEFS_OPT_INTR) + flags = ORANGEFS_OP_INTERRUPTIBLE; + + ret = service_operation(new_op, "orangefs_statfs", flags); + + if (new_op->downcall.status < 0) + goto out_op_release; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "%s: got %ld blocks available | " + "%ld blocks total | %ld block size | " + "%ld files total | %ld files avail\n", + __func__, + (long)new_op->downcall.resp.statfs.blocks_avail, + (long)new_op->downcall.resp.statfs.blocks_total, + (long)new_op->downcall.resp.statfs.block_size, + (long)new_op->downcall.resp.statfs.files_total, + (long)new_op->downcall.resp.statfs.files_avail); + + buf->f_type = sb->s_magic; + memcpy(&buf->f_fsid, &ORANGEFS_SB(sb)->fs_id, sizeof(buf->f_fsid)); + buf->f_bsize = new_op->downcall.resp.statfs.block_size; + buf->f_namelen = ORANGEFS_NAME_MAX; + + buf->f_blocks = (sector_t) new_op->downcall.resp.statfs.blocks_total; + buf->f_bfree = (sector_t) new_op->downcall.resp.statfs.blocks_avail; + buf->f_bavail = (sector_t) new_op->downcall.resp.statfs.blocks_avail; + buf->f_files = (sector_t) new_op->downcall.resp.statfs.files_total; + buf->f_ffree = (sector_t) new_op->downcall.resp.statfs.files_avail; + buf->f_frsize = sb->s_blocksize; + +out_op_release: + op_release(new_op); + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_statfs: returning %d\n", ret); + return ret; +} + +/* + * Remount as initiated by VFS layer. We just need to reparse the mount + * options, no need to signal pvfs2-client-core about it. + */ +static int orangefs_remount_fs(struct super_block *sb, int *flags, char *data) +{ + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_remount_fs: called\n"); + return parse_mount_options(sb, data, 1); +} + +/* + * Remount as initiated by pvfs2-client-core on restart. This is used to + * repopulate mount information left from previous pvfs2-client-core. + * + * the idea here is that given a valid superblock, we're + * re-initializing the user space client with the initial mount + * information specified when the super block was first initialized. + * this is very different than the first initialization/creation of a + * superblock. we use the special service_priority_operation to make + * sure that the mount gets ahead of any other pending operation that + * is waiting for servicing. this means that the pvfs2-client won't + * fail to start several times for all other pending operations before + * the client regains all of the mount information from us. + * NOTE: this function assumes that the request_mutex is already acquired! + */ +int orangefs_remount(struct orangefs_sb_info_s *orangefs_sb) +{ + struct orangefs_kernel_op_s *new_op; + int ret = -EINVAL; + + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_remount: called\n"); + + new_op = op_alloc(ORANGEFS_VFS_OP_FS_MOUNT); + if (!new_op) + return -ENOMEM; + strncpy(new_op->upcall.req.fs_mount.orangefs_config_server, + orangefs_sb->devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Attempting ORANGEFS Remount via host %s\n", + new_op->upcall.req.fs_mount.orangefs_config_server); + + /* + * we assume that the calling function has already acquired the + * request_mutex to prevent other operations from bypassing + * this one + */ + ret = service_operation(new_op, "orangefs_remount", + ORANGEFS_OP_PRIORITY | ORANGEFS_OP_NO_MUTEX); + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_remount: mount got return value of %d\n", + ret); + if (ret == 0) { + /* + * store the id assigned to this sb -- it's just a + * short-lived mapping that the system interface uses + * to map this superblock to a particular mount entry + */ + orangefs_sb->id = new_op->downcall.resp.fs_mount.id; + orangefs_sb->mount_pending = 0; + } + + op_release(new_op); + return ret; +} + +int fsid_key_table_initialize(void) +{ + return 0; +} + +void fsid_key_table_finalize(void) +{ +} + +/* Called whenever the VFS dirties the inode in response to atime updates */ +static void orangefs_dirty_inode(struct inode *inode, int flags) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_dirty_inode: %pU\n", + get_khandle_from_ino(inode)); + SetAtimeFlag(orangefs_inode); +} + +static const struct super_operations orangefs_s_ops = { + .alloc_inode = orangefs_alloc_inode, + .destroy_inode = orangefs_destroy_inode, + .dirty_inode = orangefs_dirty_inode, + .drop_inode = generic_delete_inode, + .statfs = orangefs_statfs, + .remount_fs = orangefs_remount_fs, + .show_options = generic_show_options, +}; + +static struct dentry *orangefs_fh_to_dentry(struct super_block *sb, + struct fid *fid, + int fh_len, + int fh_type) +{ + struct orangefs_object_kref refn; + + if (fh_len < 5 || fh_type > 2) + return NULL; + + ORANGEFS_khandle_from(&(refn.khandle), fid->raw, 16); + refn.fs_id = (u32) fid->raw[4]; + gossip_debug(GOSSIP_SUPER_DEBUG, + "fh_to_dentry: handle %pU, fs_id %d\n", + &refn.khandle, + refn.fs_id); + + return d_obtain_alias(orangefs_iget(sb, &refn)); +} + +static int orangefs_encode_fh(struct inode *inode, + __u32 *fh, + int *max_len, + struct inode *parent) +{ + int len = parent ? 10 : 5; + int type = 1; + struct orangefs_object_kref refn; + + if (*max_len < len) { + gossip_lerr("fh buffer is too small for encoding\n"); + *max_len = len; + type = 255; + goto out; + } + + refn = ORANGEFS_I(inode)->refn; + ORANGEFS_khandle_to(&refn.khandle, fh, 16); + fh[4] = refn.fs_id; + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Encoding fh: handle %pU, fsid %u\n", + &refn.khandle, + refn.fs_id); + + + if (parent) { + refn = ORANGEFS_I(parent)->refn; + ORANGEFS_khandle_to(&refn.khandle, (char *) fh + 20, 16); + fh[9] = refn.fs_id; + + type = 2; + gossip_debug(GOSSIP_SUPER_DEBUG, + "Encoding parent: handle %pU, fsid %u\n", + &refn.khandle, + refn.fs_id); + } + *max_len = len; + +out: + return type; +} + +static const struct export_operations orangefs_export_ops = { + .encode_fh = orangefs_encode_fh, + .fh_to_dentry = orangefs_fh_to_dentry, +}; + +static int orangefs_fill_sb(struct super_block *sb, + struct orangefs_fs_mount_response *fs_mount, + void *data, int silent) +{ + int ret = -EINVAL; + struct inode *root = NULL; + struct dentry *root_dentry = NULL; + struct orangefs_object_kref root_object; + + /* alloc and init our private orangefs sb info */ + sb->s_fs_info = kzalloc(sizeof(struct orangefs_sb_info_s), GFP_KERNEL); + if (!ORANGEFS_SB(sb)) + return -ENOMEM; + ORANGEFS_SB(sb)->sb = sb; + + ORANGEFS_SB(sb)->root_khandle = fs_mount->root_khandle; + ORANGEFS_SB(sb)->fs_id = fs_mount->fs_id; + ORANGEFS_SB(sb)->id = fs_mount->id; + + if (data) { + ret = parse_mount_options(sb, data, silent); + if (ret) + return ret; + } + + /* Hang the xattr handlers off the superblock */ + sb->s_xattr = orangefs_xattr_handlers; + sb->s_magic = ORANGEFS_SUPER_MAGIC; + sb->s_op = &orangefs_s_ops; + sb->s_d_op = &orangefs_dentry_operations; + + sb->s_blocksize = orangefs_bufmap_size_query(); + sb->s_blocksize_bits = orangefs_bufmap_shift_query(); + sb->s_maxbytes = MAX_LFS_FILESIZE; + + root_object.khandle = ORANGEFS_SB(sb)->root_khandle; + root_object.fs_id = ORANGEFS_SB(sb)->fs_id; + gossip_debug(GOSSIP_SUPER_DEBUG, + "get inode %pU, fsid %d\n", + &root_object.khandle, + root_object.fs_id); + + root = orangefs_iget(sb, &root_object); + if (IS_ERR(root)) + return PTR_ERR(root); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Allocated root inode [%p] with mode %x\n", + root, + root->i_mode); + + /* allocates and places root dentry in dcache */ + root_dentry = d_make_root(root); + if (!root_dentry) + return -ENOMEM; + + sb->s_export_op = &orangefs_export_ops; + sb->s_root = root_dentry; + return 0; +} + +struct dentry *orangefs_mount(struct file_system_type *fst, + int flags, + const char *devname, + void *data) +{ + int ret = -EINVAL; + struct super_block *sb = ERR_PTR(-EINVAL); + struct orangefs_kernel_op_s *new_op; + struct dentry *d = ERR_PTR(-EINVAL); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_mount: called with devname %s\n", + devname); + + if (!devname) { + gossip_err("ERROR: device name not specified.\n"); + return ERR_PTR(-EINVAL); + } + + new_op = op_alloc(ORANGEFS_VFS_OP_FS_MOUNT); + if (!new_op) + return ERR_PTR(-ENOMEM); + + strncpy(new_op->upcall.req.fs_mount.orangefs_config_server, + devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + gossip_debug(GOSSIP_SUPER_DEBUG, + "Attempting ORANGEFS Mount via host %s\n", + new_op->upcall.req.fs_mount.orangefs_config_server); + + ret = service_operation(new_op, "orangefs_mount", 0); + gossip_debug(GOSSIP_SUPER_DEBUG, + "orangefs_mount: mount got return value of %d\n", ret); + if (ret) + goto free_op; + + if (new_op->downcall.resp.fs_mount.fs_id == ORANGEFS_FS_ID_NULL) { + gossip_err("ERROR: Retrieved null fs_id\n"); + ret = -EINVAL; + goto free_op; + } + + sb = sget(fst, NULL, set_anon_super, flags, NULL); + + if (IS_ERR(sb)) { + d = ERR_CAST(sb); + goto free_op; + } + + ret = orangefs_fill_sb(sb, + &new_op->downcall.resp.fs_mount, data, + flags & MS_SILENT ? 1 : 0); + + if (ret) { + d = ERR_PTR(ret); + goto free_op; + } + + /* + * on successful mount, store the devname and data + * used + */ + strncpy(ORANGEFS_SB(sb)->devname, + devname, + ORANGEFS_MAX_SERVER_ADDR_LEN); + + /* mount_pending must be cleared */ + ORANGEFS_SB(sb)->mount_pending = 0; + + /* + * finally, add this sb to our list of known orangefs + * sb's + */ + gossip_debug(GOSSIP_SUPER_DEBUG, + "Adding SB %p to orangefs superblocks\n", + ORANGEFS_SB(sb)); + spin_lock(&orangefs_superblocks_lock); + list_add_tail(&ORANGEFS_SB(sb)->list, &orangefs_superblocks); + spin_unlock(&orangefs_superblocks_lock); + op_release(new_op); + return dget(sb->s_root); + +free_op: + gossip_err("orangefs_mount: mount request failed with %d\n", ret); + if (ret == -EINVAL) { + gossip_err("Ensure that all orangefs-servers have the same FS configuration files\n"); + gossip_err("Look at pvfs2-client-core log file (typically /tmp/pvfs2-client.log) for more details\n"); + } + + op_release(new_op); + + return d; +} + +void orangefs_kill_sb(struct super_block *sb) +{ + gossip_debug(GOSSIP_SUPER_DEBUG, "orangefs_kill_sb: called\n"); + + /* provided sb cleanup */ + kill_anon_super(sb); + + /* + * issue the unmount to userspace to tell it to remove the + * dynamic mount info it has for this superblock + */ + orangefs_unmount_sb(sb); + + /* remove the sb from our list of orangefs specific sb's */ + + spin_lock(&orangefs_superblocks_lock); + __list_del_entry(&ORANGEFS_SB(sb)->list); /* not list_del_init */ + ORANGEFS_SB(sb)->list.prev = NULL; + spin_unlock(&orangefs_superblocks_lock); + + /* + * make sure that ORANGEFS_DEV_REMOUNT_ALL loop that might've seen us + * gets completed before we free the dang thing. + */ + mutex_lock(&request_mutex); + mutex_unlock(&request_mutex); + + /* free the orangefs superblock private data */ + kfree(ORANGEFS_SB(sb)); +} + +int orangefs_inode_cache_initialize(void) +{ + orangefs_inode_cache = kmem_cache_create("orangefs_inode_cache", + sizeof(struct orangefs_inode_s), + 0, + ORANGEFS_CACHE_CREATE_FLAGS, + orangefs_inode_cache_ctor); + + if (!orangefs_inode_cache) { + gossip_err("Cannot create orangefs_inode_cache\n"); + return -ENOMEM; + } + return 0; +} + +int orangefs_inode_cache_finalize(void) +{ + kmem_cache_destroy(orangefs_inode_cache); + return 0; +} diff --git a/fs/orangefs/symlink.c b/fs/orangefs/symlink.c new file mode 100644 index 000000000000..6418dd638680 --- /dev/null +++ b/fs/orangefs/symlink.c @@ -0,0 +1,19 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +struct inode_operations orangefs_symlink_inode_operations = { + .readlink = generic_readlink, + .get_link = simple_get_link, + .setattr = orangefs_setattr, + .getattr = orangefs_getattr, + .listxattr = orangefs_listxattr, + .setxattr = generic_setxattr, + .permission = orangefs_permission, +}; diff --git a/fs/orangefs/upcall.h b/fs/orangefs/upcall.h new file mode 100644 index 000000000000..001b20239407 --- /dev/null +++ b/fs/orangefs/upcall.h @@ -0,0 +1,246 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +#ifndef __UPCALL_H +#define __UPCALL_H + +/* + * Sanitized this header file to fix + * 32-64 bit interaction issues between + * client-core and device + */ +struct orangefs_io_request_s { + __s32 __pad1; + __s32 buf_index; + __s32 count; + __s32 __pad2; + __s64 offset; + struct orangefs_object_kref refn; + enum ORANGEFS_io_type io_type; + __s32 readahead_size; +}; + +struct orangefs_lookup_request_s { + __s32 sym_follow; + __s32 __pad1; + struct orangefs_object_kref parent_refn; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_create_request_s { + struct orangefs_object_kref parent_refn; + struct ORANGEFS_sys_attr_s attributes; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_symlink_request_s { + struct orangefs_object_kref parent_refn; + struct ORANGEFS_sys_attr_s attributes; + char entry_name[ORANGEFS_NAME_MAX]; + char target[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_getattr_request_s { + struct orangefs_object_kref refn; + __u32 mask; + __u32 __pad1; +}; + +struct orangefs_setattr_request_s { + struct orangefs_object_kref refn; + struct ORANGEFS_sys_attr_s attributes; +}; + +struct orangefs_remove_request_s { + struct orangefs_object_kref parent_refn; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_mkdir_request_s { + struct orangefs_object_kref parent_refn; + struct ORANGEFS_sys_attr_s attributes; + char d_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_readdir_request_s { + struct orangefs_object_kref refn; + __u64 token; + __s32 max_dirent_count; + __s32 buf_index; +}; + +struct orangefs_readdirplus_request_s { + struct orangefs_object_kref refn; + __u64 token; + __s32 max_dirent_count; + __u32 mask; + __s32 buf_index; + __s32 __pad1; +}; + +struct orangefs_rename_request_s { + struct orangefs_object_kref old_parent_refn; + struct orangefs_object_kref new_parent_refn; + char d_old_name[ORANGEFS_NAME_MAX]; + char d_new_name[ORANGEFS_NAME_MAX]; +}; + +struct orangefs_statfs_request_s { + __s32 fs_id; + __s32 __pad1; +}; + +struct orangefs_truncate_request_s { + struct orangefs_object_kref refn; + __s64 size; +}; + +struct orangefs_mmap_ra_cache_flush_request_s { + struct orangefs_object_kref refn; +}; + +struct orangefs_fs_mount_request_s { + char orangefs_config_server[ORANGEFS_MAX_SERVER_ADDR_LEN]; +}; + +struct orangefs_fs_umount_request_s { + __s32 id; + __s32 fs_id; + char orangefs_config_server[ORANGEFS_MAX_SERVER_ADDR_LEN]; +}; + +struct orangefs_getxattr_request_s { + struct orangefs_object_kref refn; + __s32 key_sz; + __s32 __pad1; + char key[ORANGEFS_MAX_XATTR_NAMELEN]; +}; + +struct orangefs_setxattr_request_s { + struct orangefs_object_kref refn; + struct ORANGEFS_keyval_pair keyval; + __s32 flags; + __s32 __pad1; +}; + +struct orangefs_listxattr_request_s { + struct orangefs_object_kref refn; + __s32 requested_count; + __s32 __pad1; + __u64 token; +}; + +struct orangefs_removexattr_request_s { + struct orangefs_object_kref refn; + __s32 key_sz; + __s32 __pad1; + char key[ORANGEFS_MAX_XATTR_NAMELEN]; +}; + +struct orangefs_op_cancel_s { + __u64 op_tag; +}; + +struct orangefs_fsync_request_s { + struct orangefs_object_kref refn; +}; + +enum orangefs_param_request_type { + ORANGEFS_PARAM_REQUEST_SET = 1, + ORANGEFS_PARAM_REQUEST_GET = 2 +}; + +enum orangefs_param_request_op { + ORANGEFS_PARAM_REQUEST_OP_ACACHE_TIMEOUT_MSECS = 1, + ORANGEFS_PARAM_REQUEST_OP_ACACHE_HARD_LIMIT = 2, + ORANGEFS_PARAM_REQUEST_OP_ACACHE_SOFT_LIMIT = 3, + ORANGEFS_PARAM_REQUEST_OP_ACACHE_RECLAIM_PERCENTAGE = 4, + ORANGEFS_PARAM_REQUEST_OP_PERF_TIME_INTERVAL_SECS = 5, + ORANGEFS_PARAM_REQUEST_OP_PERF_HISTORY_SIZE = 6, + ORANGEFS_PARAM_REQUEST_OP_PERF_RESET = 7, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_TIMEOUT_MSECS = 8, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_HARD_LIMIT = 9, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_SOFT_LIMIT = 10, + ORANGEFS_PARAM_REQUEST_OP_NCACHE_RECLAIM_PERCENTAGE = 11, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_TIMEOUT_MSECS = 12, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_HARD_LIMIT = 13, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_SOFT_LIMIT = 14, + ORANGEFS_PARAM_REQUEST_OP_STATIC_ACACHE_RECLAIM_PERCENTAGE = 15, + ORANGEFS_PARAM_REQUEST_OP_CLIENT_DEBUG = 16, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_TIMEOUT_SECS = 17, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_HARD_LIMIT = 18, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_SOFT_LIMIT = 19, + ORANGEFS_PARAM_REQUEST_OP_CCACHE_RECLAIM_PERCENTAGE = 20, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_TIMEOUT_SECS = 21, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_HARD_LIMIT = 22, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_SOFT_LIMIT = 23, + ORANGEFS_PARAM_REQUEST_OP_CAPCACHE_RECLAIM_PERCENTAGE = 24, + ORANGEFS_PARAM_REQUEST_OP_TWO_MASK_VALUES = 25, +}; + +struct orangefs_param_request_s { + enum orangefs_param_request_type type; + enum orangefs_param_request_op op; + __s64 value; + char s_value[ORANGEFS_MAX_DEBUG_STRING_LEN]; +}; + +enum orangefs_perf_count_request_type { + ORANGEFS_PERF_COUNT_REQUEST_ACACHE = 1, + ORANGEFS_PERF_COUNT_REQUEST_NCACHE = 2, + ORANGEFS_PERF_COUNT_REQUEST_CAPCACHE = 3, +}; + +struct orangefs_perf_count_request_s { + enum orangefs_perf_count_request_type type; + __s32 __pad1; +}; + +struct orangefs_fs_key_request_s { + __s32 fsid; + __s32 __pad1; +}; + +struct orangefs_upcall_s { + __s32 type; + __u32 uid; + __u32 gid; + int pid; + int tgid; + /* Trailers unused but must be retained for protocol compatibility. */ + __s64 trailer_size; + char *trailer_buf; + + union { + struct orangefs_io_request_s io; + struct orangefs_lookup_request_s lookup; + struct orangefs_create_request_s create; + struct orangefs_symlink_request_s sym; + struct orangefs_getattr_request_s getattr; + struct orangefs_setattr_request_s setattr; + struct orangefs_remove_request_s remove; + struct orangefs_mkdir_request_s mkdir; + struct orangefs_readdir_request_s readdir; + struct orangefs_readdirplus_request_s readdirplus; + struct orangefs_rename_request_s rename; + struct orangefs_statfs_request_s statfs; + struct orangefs_truncate_request_s truncate; + struct orangefs_mmap_ra_cache_flush_request_s ra_cache_flush; + struct orangefs_fs_mount_request_s fs_mount; + struct orangefs_fs_umount_request_s fs_umount; + struct orangefs_getxattr_request_s getxattr; + struct orangefs_setxattr_request_s setxattr; + struct orangefs_listxattr_request_s listxattr; + struct orangefs_removexattr_request_s removexattr; + struct orangefs_op_cancel_s cancel; + struct orangefs_fsync_request_s fsync; + struct orangefs_param_request_s param; + struct orangefs_perf_count_request_s perf_count; + struct orangefs_fs_key_request_s fs_key; + } req; +}; + +#endif /* __UPCALL_H */ diff --git a/fs/orangefs/waitqueue.c b/fs/orangefs/waitqueue.c new file mode 100644 index 000000000000..31635bc303fe --- /dev/null +++ b/fs/orangefs/waitqueue.c @@ -0,0 +1,357 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * (C) 2011 Omnibond Systems + * + * Changes by Acxiom Corporation to implement generic service_operation() + * function, Copyright Acxiom Corporation, 2005. + * + * See COPYING in top-level directory. + */ + +/* + * In-kernel waitqueue operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" + +static int wait_for_matching_downcall(struct orangefs_kernel_op_s *, long, bool); +static void orangefs_clean_up_interrupted_operation(struct orangefs_kernel_op_s *); + +/* + * What we do in this function is to walk the list of operations that are + * present in the request queue and mark them as purged. + * NOTE: This is called from the device close after client-core has + * guaranteed that no new operations could appear on the list since the + * client-core is anyway going to exit. + */ +void purge_waiting_ops(void) +{ + struct orangefs_kernel_op_s *op; + + spin_lock(&orangefs_request_list_lock); + list_for_each_entry(op, &orangefs_request_list, list) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "pvfs2-client-core: purging op tag %llu %s\n", + llu(op->tag), + get_opname_string(op)); + set_op_state_purged(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + } + spin_unlock(&orangefs_request_list_lock); +} + +/* + * submits a ORANGEFS operation and waits for it to complete + * + * Note op->downcall.status will contain the status of the operation (in + * errno format), whether provided by pvfs2-client or a result of failure to + * service the operation. If the caller wishes to distinguish, then + * op->state can be checked to see if it was serviced or not. + * + * Returns contents of op->downcall.status for convenience + */ +int service_operation(struct orangefs_kernel_op_s *op, + const char *op_name, + int flags) +{ + long timeout = MAX_SCHEDULE_TIMEOUT; + int ret = 0; + + DEFINE_WAIT(wait_entry); + + op->upcall.tgid = current->tgid; + op->upcall.pid = current->pid; + +retry_servicing: + op->downcall.status = 0; + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: %s op:%p: process:%s: pid:%d:\n", + __func__, + op_name, + op, + current->comm, + current->pid); + + /* + * If ORANGEFS_OP_NO_MUTEX was set in flags, we need to avoid + * acquiring the request_mutex because we're servicing a + * high priority remount operation and the request_mutex is + * already taken. + */ + if (!(flags & ORANGEFS_OP_NO_MUTEX)) { + if (flags & ORANGEFS_OP_INTERRUPTIBLE) + ret = mutex_lock_interruptible(&request_mutex); + else + ret = mutex_lock_killable(&request_mutex); + /* + * check to see if we were interrupted while waiting for + * mutex + */ + if (ret < 0) { + op->downcall.status = ret; + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: service_operation interrupted.\n", + __func__); + return ret; + } + } + + /* queue up the operation */ + spin_lock(&orangefs_request_list_lock); + spin_lock(&op->lock); + set_op_state_waiting(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + /* add high priority remount op to the front of the line. */ + if (flags & ORANGEFS_OP_PRIORITY) + list_add(&op->list, &orangefs_request_list); + else + list_add_tail(&op->list, &orangefs_request_list); + spin_unlock(&op->lock); + wake_up_interruptible(&orangefs_request_list_waitq); + if (!__is_daemon_in_service()) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s:client core is NOT in service.\n", + __func__); + timeout = op_timeout_secs * HZ; + } + spin_unlock(&orangefs_request_list_lock); + + if (!(flags & ORANGEFS_OP_NO_MUTEX)) + mutex_unlock(&request_mutex); + + ret = wait_for_matching_downcall(op, timeout, + flags & ORANGEFS_OP_INTERRUPTIBLE); + + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: wait_for_matching_downcall returned %d for %p\n", + __func__, + ret, + op); + + /* got matching downcall; make sure status is in errno format */ + if (!ret) { + spin_unlock(&op->lock); + op->downcall.status = + orangefs_normalize_to_errno(op->downcall.status); + ret = op->downcall.status; + goto out; + } + + /* failed to get matching downcall */ + if (ret == -ETIMEDOUT) { + gossip_err("%s: %s -- wait timed out; aborting attempt.\n", + __func__, + op_name); + } + + /* + * remove a waiting op from the request list or + * remove an in-progress op from the in-progress list. + */ + orangefs_clean_up_interrupted_operation(op); + + op->downcall.status = ret; + /* retry if operation has not been serviced and if requested */ + if (ret == -EAGAIN) { + op->attempts++; + timeout = op_timeout_secs * HZ; + gossip_debug(GOSSIP_WAIT_DEBUG, + "orangefs: tag %llu (%s)" + " -- operation to be retried (%d attempt)\n", + llu(op->tag), + op_name, + op->attempts); + + /* + * io ops (ops that use the shared memory buffer) have + * to be returned to their caller for a retry. Other ops + * can just be recycled here. + */ + if (!op->uses_shared_memory) + goto retry_servicing; + } + +out: + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: %s returning: %d for %p.\n", + __func__, + op_name, + ret, + op); + return ret; +} + +/* This can get called on an I/O op if it had a bad service_operation. */ +bool orangefs_cancel_op_in_progress(struct orangefs_kernel_op_s *op) +{ + u64 tag = op->tag; + if (!op_state_in_progress(op)) + return false; + + op->slot_to_free = op->upcall.req.io.buf_index; + memset(&op->upcall, 0, sizeof(op->upcall)); + memset(&op->downcall, 0, sizeof(op->downcall)); + op->upcall.type = ORANGEFS_VFS_OP_CANCEL; + op->upcall.req.cancel.op_tag = tag; + op->downcall.type = ORANGEFS_VFS_OP_INVALID; + op->downcall.status = -1; + orangefs_new_tag(op); + + spin_lock(&orangefs_request_list_lock); + /* orangefs_request_list_lock is enough of a barrier here */ + if (!__is_daemon_in_service()) { + spin_unlock(&orangefs_request_list_lock); + return false; + } + spin_lock(&op->lock); + set_op_state_waiting(op); + gossip_debug(GOSSIP_DEV_DEBUG, + "%s: op:%s: op_state:%d: process:%s:\n", + __func__, + get_opname_string(op), + op->op_state, + current->comm); + list_add(&op->list, &orangefs_request_list); + spin_unlock(&op->lock); + spin_unlock(&orangefs_request_list_lock); + + gossip_debug(GOSSIP_WAIT_DEBUG, + "Attempting ORANGEFS operation cancellation of tag %llu\n", + llu(tag)); + return true; +} + +/* + * Change an op to the "given up" state and remove it from its list. + */ +static void + orangefs_clean_up_interrupted_operation(struct orangefs_kernel_op_s *op) +{ + /* + * handle interrupted cases depending on what state we were in when + * the interruption is detected. + * + * Called with op->lock held. + */ + + /* + * List manipulation code elsewhere will ignore ops that + * have been given up upon. + */ + op->op_state |= OP_VFS_STATE_GIVEN_UP; + + if (list_empty(&op->list)) { + /* caught copying to/from daemon */ + BUG_ON(op_state_serviced(op)); + spin_unlock(&op->lock); + wait_for_completion(&op->waitq); + } else if (op_state_waiting(op)) { + /* + * upcall hasn't been read; remove op from upcall request + * list. + */ + spin_unlock(&op->lock); + spin_lock(&orangefs_request_list_lock); + list_del_init(&op->list); + spin_unlock(&orangefs_request_list_lock); + gossip_debug(GOSSIP_WAIT_DEBUG, + "Interrupted: Removed op %p from request_list\n", + op); + } else if (op_state_in_progress(op)) { + /* op must be removed from the in progress htable */ + spin_unlock(&op->lock); + spin_lock(&htable_ops_in_progress_lock); + list_del_init(&op->list); + spin_unlock(&htable_ops_in_progress_lock); + gossip_debug(GOSSIP_WAIT_DEBUG, + "Interrupted: Removed op %p" + " from htable_ops_in_progress\n", + op); + } else { + spin_unlock(&op->lock); + gossip_err("interrupted operation is in a weird state 0x%x\n", + op->op_state); + } + reinit_completion(&op->waitq); +} + +/* + * Sleeps on waitqueue waiting for matching downcall. + * If client-core finishes servicing, then we are good to go. + * else if client-core exits, we get woken up here, and retry with a timeout + * + * When this call returns to the caller, the specified op will no + * longer be in either the in_progress hash table or on the request list. + * + * Returns 0 on success and -errno on failure + * Errors are: + * EAGAIN in case we want the caller to requeue and try again.. + * EINTR/EIO/ETIMEDOUT indicating we are done trying to service this + * operation since client-core seems to be exiting too often + * or if we were interrupted. + * + * Returns with op->lock taken. + */ +static int wait_for_matching_downcall(struct orangefs_kernel_op_s *op, + long timeout, + bool interruptible) +{ + long n; + + /* + * There's a "schedule_timeout" inside of these wait + * primitives, during which the op is out of the hands of the + * user process that needs something done and is being + * manipulated by the client-core process. + */ + if (interruptible) + n = wait_for_completion_interruptible_timeout(&op->waitq, + timeout); + else + n = wait_for_completion_killable_timeout(&op->waitq, timeout); + + spin_lock(&op->lock); + + if (op_state_serviced(op)) + return 0; + + if (unlikely(n < 0)) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: operation interrupted, tag %llu, %p\n", + __func__, + llu(op->tag), + op); + return -EINTR; + } + if (op_state_purged(op)) { + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: operation purged, tag %llu, %p, %d\n", + __func__, + llu(op->tag), + op, + op->attempts); + return (op->attempts < ORANGEFS_PURGE_RETRY_COUNT) ? + -EAGAIN : + -EIO; + } + /* must have timed out, then... */ + gossip_debug(GOSSIP_WAIT_DEBUG, + "%s: operation timed out, tag %llu, %p, %d)\n", + __func__, + llu(op->tag), + op, + op->attempts); + return -ETIMEDOUT; +} diff --git a/fs/orangefs/xattr.c b/fs/orangefs/xattr.c new file mode 100644 index 000000000000..ef5da7538cd5 --- /dev/null +++ b/fs/orangefs/xattr.c @@ -0,0 +1,545 @@ +/* + * (C) 2001 Clemson University and The University of Chicago + * + * See COPYING in top-level directory. + */ + +/* + * Linux VFS extended attribute operations. + */ + +#include "protocol.h" +#include "orangefs-kernel.h" +#include "orangefs-bufmap.h" +#include <linux/posix_acl_xattr.h> +#include <linux/xattr.h> + + +#define SYSTEM_ORANGEFS_KEY "system.pvfs2." +#define SYSTEM_ORANGEFS_KEY_LEN 13 + +/* + * this function returns + * 0 if the key corresponding to name is not meant to be printed as part + * of a listxattr. + * 1 if the key corresponding to name is meant to be returned as part of + * a listxattr. + * The ones that start SYSTEM_ORANGEFS_KEY are the ones to avoid printing. + */ +static int is_reserved_key(const char *key, size_t size) +{ + + if (size < SYSTEM_ORANGEFS_KEY_LEN) + return 1; + + return strncmp(key, SYSTEM_ORANGEFS_KEY, SYSTEM_ORANGEFS_KEY_LEN) ? 1 : 0; +} + +static inline int convert_to_internal_xattr_flags(int setxattr_flags) +{ + int internal_flag = 0; + + if (setxattr_flags & XATTR_REPLACE) { + /* Attribute must exist! */ + internal_flag = ORANGEFS_XATTR_REPLACE; + } else if (setxattr_flags & XATTR_CREATE) { + /* Attribute must not exist */ + internal_flag = ORANGEFS_XATTR_CREATE; + } + return internal_flag; +} + + +/* + * Tries to get a specified key's attributes of a given + * file into a user-specified buffer. Note that the getxattr + * interface allows for the users to probe the size of an + * extended attribute by passing in a value of 0 to size. + * Thus our return value is always the size of the attribute + * unless the key does not exist for the file and/or if + * there were errors in fetching the attribute value. + */ +ssize_t orangefs_inode_getxattr(struct inode *inode, const char *prefix, + const char *name, void *buffer, size_t size) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op = NULL; + ssize_t ret = -ENOMEM; + ssize_t length = 0; + int fsuid; + int fsgid; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "%s: prefix %s name %s, buffer_size %zd\n", + __func__, prefix, name, size); + + if (name == NULL || (size > 0 && buffer == NULL)) { + gossip_err("orangefs_inode_getxattr: bogus NULL pointers\n"); + return -EINVAL; + } + if ((strlen(name) + strlen(prefix)) >= ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err("Invalid key length (%d)\n", + (int)(strlen(name) + strlen(prefix))); + return -EINVAL; + } + + fsuid = from_kuid(current_user_ns(), current_fsuid()); + fsgid = from_kgid(current_user_ns(), current_fsgid()); + + gossip_debug(GOSSIP_XATTR_DEBUG, + "getxattr on inode %pU, name %s " + "(uid %o, gid %o)\n", + get_khandle_from_ino(inode), + name, + fsuid, + fsgid); + + down_read(&orangefs_inode->xattr_sem); + + new_op = op_alloc(ORANGEFS_VFS_OP_GETXATTR); + if (!new_op) + goto out_unlock; + + new_op->upcall.req.getxattr.refn = orangefs_inode->refn; + ret = snprintf((char *)new_op->upcall.req.getxattr.key, + ORANGEFS_MAX_XATTR_NAMELEN, "%s%s", prefix, name); + + /* + * NOTE: Although keys are meant to be NULL terminated textual + * strings, I am going to explicitly pass the length just in case + * we change this later on... + */ + new_op->upcall.req.getxattr.key_sz = ret + 1; + + ret = service_operation(new_op, "orangefs_inode_getxattr", + get_interruptible_flag(inode)); + if (ret != 0) { + if (ret == -ENOENT) { + ret = -ENODATA; + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_getxattr: inode %pU key %s" + " does not exist!\n", + get_khandle_from_ino(inode), + (char *)new_op->upcall.req.getxattr.key); + } + goto out_release_op; + } + + /* + * Length returned includes null terminator. + */ + length = new_op->downcall.resp.getxattr.val_sz; + + /* + * Just return the length of the queried attribute. + */ + if (size == 0) { + ret = length; + goto out_release_op; + } + + /* + * Check to see if key length is > provided buffer size. + */ + if (length > size) { + ret = -ERANGE; + goto out_release_op; + } + + memset(buffer, 0, size); + memcpy(buffer, new_op->downcall.resp.getxattr.val, length); + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_getxattr: inode %pU " + "key %s key_sz %d, val_len %d\n", + get_khandle_from_ino(inode), + (char *)new_op-> + upcall.req.getxattr.key, + (int)new_op-> + upcall.req.getxattr.key_sz, + (int)ret); + + ret = length; + +out_release_op: + op_release(new_op); +out_unlock: + up_read(&orangefs_inode->xattr_sem); + return ret; +} + +static int orangefs_inode_removexattr(struct inode *inode, + const char *prefix, + const char *name, + int flags) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op = NULL; + int ret = -ENOMEM; + + down_write(&orangefs_inode->xattr_sem); + new_op = op_alloc(ORANGEFS_VFS_OP_REMOVEXATTR); + if (!new_op) + goto out_unlock; + + new_op->upcall.req.removexattr.refn = orangefs_inode->refn; + /* + * NOTE: Although keys are meant to be NULL terminated + * textual strings, I am going to explicitly pass the + * length just in case we change this later on... + */ + ret = snprintf((char *)new_op->upcall.req.removexattr.key, + ORANGEFS_MAX_XATTR_NAMELEN, + "%s%s", + (prefix ? prefix : ""), + name); + new_op->upcall.req.removexattr.key_sz = ret + 1; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_removexattr: key %s, key_sz %d\n", + (char *)new_op->upcall.req.removexattr.key, + (int)new_op->upcall.req.removexattr.key_sz); + + ret = service_operation(new_op, + "orangefs_inode_removexattr", + get_interruptible_flag(inode)); + if (ret == -ENOENT) { + /* + * Request to replace a non-existent attribute is an error. + */ + if (flags & XATTR_REPLACE) + ret = -ENODATA; + else + ret = 0; + } + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_removexattr: returning %d\n", ret); + + op_release(new_op); +out_unlock: + up_write(&orangefs_inode->xattr_sem); + return ret; +} + +/* + * Tries to set an attribute for a given key on a file. + * + * Returns a -ve number on error and 0 on success. Key is text, but value + * can be binary! + */ +int orangefs_inode_setxattr(struct inode *inode, const char *prefix, + const char *name, const void *value, size_t size, int flags) +{ + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + int internal_flag = 0; + int ret = -ENOMEM; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "%s: prefix %s, name %s, buffer_size %zd\n", + __func__, prefix, name, size); + + if (size < 0 || + size >= ORANGEFS_MAX_XATTR_VALUELEN || + flags < 0) { + gossip_err("orangefs_inode_setxattr: bogus values of size(%d), flags(%d)\n", + (int)size, + flags); + return -EINVAL; + } + + if (name == NULL || + (size > 0 && value == NULL)) { + gossip_err("orangefs_inode_setxattr: bogus NULL pointers!\n"); + return -EINVAL; + } + + internal_flag = convert_to_internal_xattr_flags(flags); + + if (prefix) { + if (strlen(name) + strlen(prefix) >= ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err + ("orangefs_inode_setxattr: bogus key size (%d)\n", + (int)(strlen(name) + strlen(prefix))); + return -EINVAL; + } + } else { + if (strlen(name) >= ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err + ("orangefs_inode_setxattr: bogus key size (%d)\n", + (int)(strlen(name))); + return -EINVAL; + } + } + + /* This is equivalent to a removexattr */ + if (size == 0 && value == NULL) { + gossip_debug(GOSSIP_XATTR_DEBUG, + "removing xattr (%s%s)\n", + prefix, + name); + return orangefs_inode_removexattr(inode, prefix, name, flags); + } + + gossip_debug(GOSSIP_XATTR_DEBUG, + "setxattr on inode %pU, name %s\n", + get_khandle_from_ino(inode), + name); + + down_write(&orangefs_inode->xattr_sem); + new_op = op_alloc(ORANGEFS_VFS_OP_SETXATTR); + if (!new_op) + goto out_unlock; + + + new_op->upcall.req.setxattr.refn = orangefs_inode->refn; + new_op->upcall.req.setxattr.flags = internal_flag; + /* + * NOTE: Although keys are meant to be NULL terminated textual + * strings, I am going to explicitly pass the length just in + * case we change this later on... + */ + ret = snprintf((char *)new_op->upcall.req.setxattr.keyval.key, + ORANGEFS_MAX_XATTR_NAMELEN, + "%s%s", + prefix, name); + new_op->upcall.req.setxattr.keyval.key_sz = ret + 1; + memcpy(new_op->upcall.req.setxattr.keyval.val, value, size); + new_op->upcall.req.setxattr.keyval.val_sz = size; + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_setxattr: key %s, key_sz %d " + " value size %zd\n", + (char *)new_op->upcall.req.setxattr.keyval.key, + (int)new_op->upcall.req.setxattr.keyval.key_sz, + size); + + ret = service_operation(new_op, + "orangefs_inode_setxattr", + get_interruptible_flag(inode)); + + gossip_debug(GOSSIP_XATTR_DEBUG, + "orangefs_inode_setxattr: returning %d\n", + ret); + + /* when request is serviced properly, free req op struct */ + op_release(new_op); +out_unlock: + up_write(&orangefs_inode->xattr_sem); + return ret; +} + +/* + * Tries to get a specified object's keys into a user-specified buffer of a + * given size. Note that like the previous instances of xattr routines, this + * also allows you to pass in a NULL pointer and 0 size to probe the size for + * subsequent memory allocations. Thus our return value is always the size of + * all the keys unless there were errors in fetching the keys! + */ +ssize_t orangefs_listxattr(struct dentry *dentry, char *buffer, size_t size) +{ + struct inode *inode = dentry->d_inode; + struct orangefs_inode_s *orangefs_inode = ORANGEFS_I(inode); + struct orangefs_kernel_op_s *new_op; + __u64 token = ORANGEFS_ITERATE_START; + ssize_t ret = -ENOMEM; + ssize_t total = 0; + int count_keys = 0; + int key_size; + int i = 0; + int returned_count = 0; + + if (size > 0 && buffer == NULL) { + gossip_err("%s: bogus NULL pointers\n", __func__); + return -EINVAL; + } + if (size < 0) { + gossip_err("Invalid size (%d)\n", (int)size); + return -EINVAL; + } + + down_read(&orangefs_inode->xattr_sem); + new_op = op_alloc(ORANGEFS_VFS_OP_LISTXATTR); + if (!new_op) + goto out_unlock; + + if (buffer && size > 0) + memset(buffer, 0, size); + +try_again: + key_size = 0; + new_op->upcall.req.listxattr.refn = orangefs_inode->refn; + new_op->upcall.req.listxattr.token = token; + new_op->upcall.req.listxattr.requested_count = + (size == 0) ? 0 : ORANGEFS_MAX_XATTR_LISTLEN; + ret = service_operation(new_op, __func__, + get_interruptible_flag(inode)); + if (ret != 0) + goto done; + + if (size == 0) { + /* + * This is a bit of a big upper limit, but I did not want to + * spend too much time getting this correct, since users end + * up allocating memory rather than us... + */ + total = new_op->downcall.resp.listxattr.returned_count * + ORANGEFS_MAX_XATTR_NAMELEN; + goto done; + } + + returned_count = new_op->downcall.resp.listxattr.returned_count; + if (returned_count < 0 || + returned_count >= ORANGEFS_MAX_XATTR_LISTLEN) { + gossip_err("%s: impossible value for returned_count:%d:\n", + __func__, + returned_count); + ret = -EIO; + goto done; + } + + /* + * Check to see how much can be fit in the buffer. Fit only whole keys. + */ + for (i = 0; i < returned_count; i++) { + if (new_op->downcall.resp.listxattr.lengths[i] < 0 || + new_op->downcall.resp.listxattr.lengths[i] > + ORANGEFS_MAX_XATTR_NAMELEN) { + gossip_err("%s: impossible value for lengths[%d]\n", + __func__, + new_op->downcall.resp.listxattr.lengths[i]); + ret = -EIO; + goto done; + } + if (total + new_op->downcall.resp.listxattr.lengths[i] > size) + goto done; + + /* + * Since many dumb programs try to setxattr() on our reserved + * xattrs this is a feeble attempt at defeating those by not + * listing them in the output of listxattr.. sigh + */ + if (is_reserved_key(new_op->downcall.resp.listxattr.key + + key_size, + new_op->downcall.resp. + listxattr.lengths[i])) { + gossip_debug(GOSSIP_XATTR_DEBUG, "Copying key %d -> %s\n", + i, new_op->downcall.resp.listxattr.key + + key_size); + memcpy(buffer + total, + new_op->downcall.resp.listxattr.key + key_size, + new_op->downcall.resp.listxattr.lengths[i]); + total += new_op->downcall.resp.listxattr.lengths[i]; + count_keys++; + } else { + gossip_debug(GOSSIP_XATTR_DEBUG, "[RESERVED] key %d -> %s\n", + i, new_op->downcall.resp.listxattr.key + + key_size); + } + key_size += new_op->downcall.resp.listxattr.lengths[i]; + } + + /* + * Since the buffer was large enough, we might have to continue + * fetching more keys! + */ + token = new_op->downcall.resp.listxattr.token; + if (token != ORANGEFS_ITERATE_END) + goto try_again; + +done: + gossip_debug(GOSSIP_XATTR_DEBUG, "%s: returning %d" + " [size of buffer %ld] (filled in %d keys)\n", + __func__, + ret ? (int)ret : (int)total, + (long)size, + count_keys); + op_release(new_op); + if (ret == 0) + ret = total; +out_unlock: + up_read(&orangefs_inode->xattr_sem); + return ret; +} + +static int orangefs_xattr_set_default(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + const void *buffer, + size_t size, + int flags) +{ + return orangefs_inode_setxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + name, + buffer, + size, + flags); +} + +static int orangefs_xattr_get_default(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + void *buffer, + size_t size) +{ + return orangefs_inode_getxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + name, + buffer, + size); + +} + +static int orangefs_xattr_set_trusted(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + const void *buffer, + size_t size, + int flags) +{ + return orangefs_inode_setxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_TRUSTED_PREFIX, + name, + buffer, + size, + flags); +} + +static int orangefs_xattr_get_trusted(const struct xattr_handler *handler, + struct dentry *dentry, + const char *name, + void *buffer, + size_t size) +{ + return orangefs_inode_getxattr(dentry->d_inode, + ORANGEFS_XATTR_NAME_TRUSTED_PREFIX, + name, + buffer, + size); +} + +static struct xattr_handler orangefs_xattr_trusted_handler = { + .prefix = ORANGEFS_XATTR_NAME_TRUSTED_PREFIX, + .get = orangefs_xattr_get_trusted, + .set = orangefs_xattr_set_trusted, +}; + +static struct xattr_handler orangefs_xattr_default_handler = { + /* + * NOTE: this is set to be the empty string. + * so that all un-prefixed xattrs keys get caught + * here! + */ + .prefix = ORANGEFS_XATTR_NAME_DEFAULT_PREFIX, + .get = orangefs_xattr_get_default, + .set = orangefs_xattr_set_default, +}; + +const struct xattr_handler *orangefs_xattr_handlers[] = { + &posix_acl_access_xattr_handler, + &posix_acl_default_xattr_handler, + &orangefs_xattr_trusted_handler, + &orangefs_xattr_default_handler, + NULL +}; diff --git a/fs/ubifs/Makefile b/fs/ubifs/Makefile index 2c6f0cb816b4..c54a24360f85 100644 --- a/fs/ubifs/Makefile +++ b/fs/ubifs/Makefile @@ -4,3 +4,4 @@ ubifs-y += shrinker.o journal.o file.o dir.o super.o sb.o io.o ubifs-y += tnc.o master.o scan.o replay.o log.o commit.o gc.o orphan.o ubifs-y += budget.o find.o tnc_commit.o compress.o lpt.o lprops.o ubifs-y += recovery.o ioctl.o lpt_commit.o tnc_misc.o xattr.o debug.o +ubifs-y += misc.o diff --git a/fs/ubifs/misc.c b/fs/ubifs/misc.c new file mode 100644 index 000000000000..486a2844949f --- /dev/null +++ b/fs/ubifs/misc.c @@ -0,0 +1,57 @@ +#include <linux/kernel.h> +#include "ubifs.h" + +/* Normal UBIFS messages */ +void ubifs_msg(const struct ubifs_info *c, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + pr_notice("UBIFS (ubi%d:%d): %pV\n", + c->vi.ubi_num, c->vi.vol_id, &vaf); + + va_end(args); +} \ + +/* UBIFS error messages */ +void ubifs_err(const struct ubifs_info *c, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + pr_err("UBIFS error (ubi%d:%d pid %d): %ps: %pV\n", + c->vi.ubi_num, c->vi.vol_id, current->pid, + __builtin_return_address(0), + &vaf); + + va_end(args); +} \ + +/* UBIFS warning messages */ +void ubifs_warn(const struct ubifs_info *c, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + + vaf.fmt = fmt; + vaf.va = &args; + + pr_warn("UBIFS warning (ubi%d:%d pid %d): %ps: %pV\n", + c->vi.ubi_num, c->vi.vol_id, current->pid, + __builtin_return_address(0), + &vaf); + + va_end(args); +} diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index a5697de763f5..c2a57e193a81 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -42,30 +42,6 @@ /* Version of this UBIFS implementation */ #define UBIFS_VERSION 1 -/* Normal UBIFS messages */ -#define ubifs_msg(c, fmt, ...) \ - pr_notice("UBIFS (ubi%d:%d): " fmt "\n", \ - (c)->vi.ubi_num, (c)->vi.vol_id, ##__VA_ARGS__) -/* UBIFS error messages */ -#define ubifs_err(c, fmt, ...) \ - pr_err("UBIFS error (ubi%d:%d pid %d): %s: " fmt "\n", \ - (c)->vi.ubi_num, (c)->vi.vol_id, current->pid, \ - __func__, ##__VA_ARGS__) -/* UBIFS warning messages */ -#define ubifs_warn(c, fmt, ...) \ - pr_warn("UBIFS warning (ubi%d:%d pid %d): %s: " fmt "\n", \ - (c)->vi.ubi_num, (c)->vi.vol_id, current->pid, \ - __func__, ##__VA_ARGS__) -/* - * A variant of 'ubifs_err()' which takes the UBIFS file-sytem description - * object as an argument. - */ -#define ubifs_errc(c, fmt, ...) \ - do { \ - if (!(c)->probing) \ - ubifs_err(c, fmt, ##__VA_ARGS__); \ - } while (0) - /* UBIFS file system VFS magic number */ #define UBIFS_SUPER_MAGIC 0x24051905 @@ -1802,4 +1778,21 @@ int ubifs_decompress(const struct ubifs_info *c, const void *buf, int len, #include "misc.h" #include "key.h" +/* Normal UBIFS messages */ +__printf(2, 3) +void ubifs_msg(const struct ubifs_info *c, const char *fmt, ...); +__printf(2, 3) +void ubifs_err(const struct ubifs_info *c, const char *fmt, ...); +__printf(2, 3) +void ubifs_warn(const struct ubifs_info *c, const char *fmt, ...); +/* + * A variant of 'ubifs_err()' which takes the UBIFS file-sytem description + * object as an argument. + */ +#define ubifs_errc(c, fmt, ...) \ +do { \ + if (!(c)->probing) \ + ubifs_err(c, fmt, ##__VA_ARGS__); \ +} while (0) + #endif /* !__UBIFS_H__ */ diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c index c7f4d434d098..b043e044121d 100644 --- a/fs/ubifs/xattr.c +++ b/fs/ubifs/xattr.c @@ -59,7 +59,6 @@ #include <linux/fs.h> #include <linux/slab.h> #include <linux/xattr.h> -#include <linux/posix_acl_xattr.h> /* * Limit the number of extended attributes per inode so that the total size diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index f64639176670..3542d94fddce 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -121,4 +121,5 @@ xfs-$(CONFIG_XFS_RT) += xfs_rtalloc.o xfs-$(CONFIG_XFS_POSIX_ACL) += xfs_acl.o xfs-$(CONFIG_SYSCTL) += xfs_sysctl.o xfs-$(CONFIG_COMPAT) += xfs_ioctl32.o -xfs-$(CONFIG_NFSD_PNFS) += xfs_pnfs.o +xfs-$(CONFIG_NFSD_BLOCKLAYOUT) += xfs_pnfs.o +xfs-$(CONFIG_NFSD_SCSILAYOUT) += xfs_pnfs.o diff --git a/fs/xfs/xfs_export.c b/fs/xfs/xfs_export.c index 2816d42507bc..a1b2dd828b9d 100644 --- a/fs/xfs/xfs_export.c +++ b/fs/xfs/xfs_export.c @@ -246,7 +246,7 @@ const struct export_operations xfs_export_operations = { .fh_to_parent = xfs_fs_fh_to_parent, .get_parent = xfs_fs_get_parent, .commit_metadata = xfs_fs_nfs_commit_metadata, -#ifdef CONFIG_NFSD_PNFS +#ifdef CONFIG_NFSD_BLOCKLAYOUT .get_uuid = xfs_fs_get_uuid, .map_blocks = xfs_fs_map_blocks, .commit_blocks = xfs_fs_commit_blocks, diff --git a/fs/xfs/xfs_pnfs.h b/fs/xfs/xfs_pnfs.h index 8147ac108820..93f74853961b 100644 --- a/fs/xfs/xfs_pnfs.h +++ b/fs/xfs/xfs_pnfs.h @@ -1,7 +1,7 @@ #ifndef _XFS_PNFS_H #define _XFS_PNFS_H 1 -#ifdef CONFIG_NFSD_PNFS +#if defined(CONFIG_NFSD_BLOCKLAYOUT) || defined(CONFIG_NFSD_SCSILAYOUT) int xfs_fs_get_uuid(struct super_block *sb, u8 *buf, u32 *len, u64 *offset); int xfs_fs_map_blocks(struct inode *inode, loff_t offset, u64 length, struct iomap *iomap, bool write, u32 *device_generation); |
